Product updates

Changelog

A dated, plain-language record of what changed and where you will notice it.

Daily product release

UI 1.8.26 · Extension 1.20.6

Legal-page scope is consistent

Privacy and Terms now use the same concise description of where each document applies.

Improved

Cleaner document introduction

Both pages now identify the website, web app, Chrome extension, and service emails without an extra divider above the effective date.

Improved

A more direct legal-page layout

The summary cards and boxed contact panels have been removed from both legal pages so users can read the full document in one continuous layout.

Improved

Privacy practices are documented clearly

The Privacy Policy now explains account, analysis, extension, payment, email, security-location, provider, cookie, retention, and user-choice practices in one navigable document.

Improved

Service expectations are complete

The Terms now explain account responsibilities, analysis limits, submitted content, acceptable use, prepaid access, communications, availability, and legal protections without presenting the score as a truth verdict.

Fixed

Publication email starts immediately

Selecting email during publication now begins delivery instead of leaving the message waiting for a separate scheduled check.

New

Published items can be emailed later

A published article or announcement can be sent from its editorial page, including resuming an unfinished send without repeating completed recipients or deliberately sending a new copy.

Improved

Blog link previews reflect the current page

Links to Veritas Shield pages inside an article now preview the live page rather than a manually maintained picture that can fall out of date.

New

A welcome after account creation

New email and Google accounts receive one welcome message without callback or signup retries sending it twice.

Improved

Verification emails stay securely generated

Email codes continue to be created and protected by the verification service while using the currently approved message design.

Improved

Terms and Privacy keep distinct messages

Legal updates preserve their own heading, explanation, document link, and Veritas Shield Legal sender.

Improved

Long email copy stays visually balanced

Titles and descriptions now choose balanced word breaks before sending, keeping multi-line messages easier to scan across phones and larger screens.

New

Manage email choices in Profile

Users can subscribe or unsubscribe from articles, announcements, and product updates at any time.

Improved

Read published content inside the email

Article and announcement emails can include the cover, written content, in-article pictures, and a direct link to the website.

New

Country-change sign-in alerts

After a country has been seen for an account, a later sign-in from another country sends a security alert to the account email.

Improved

Clearer legal notices

Terms of Service and Privacy Policy messages now keep distinct names, summaries, and document links.

Improved

Clearer sender identities

Users can recognize whether an email is an article, announcement, newsletter, or product update from its sender address, with one consistent fallback when a dedicated identity is not configured.

New

Dedicated article and announcement delivery

Subscribed updates can be sent from the verified Veritas Shield news domain while verification codes and account-security messages stay separate.

Improved

Retries do not create duplicate email

Temporary provider errors and sending limits pause safely, then resume with the same protected delivery identity.

Improved

Deep Scan setup fits without scrolling

The Analyzer now gives the purpose, browser status, and one clear install action in a smaller dialog.

Improved

Profile shows the complete plan

Plan details now include unlimited analysis, Chrome Deep Scan, explained reporting scores, Reader, Evidence Atlas, comparison, and private 30-day History.

New

Sign in with an email code

Existing users can request a short-lived code instead of entering a password.

Improved

One verification path for password changes

Password recovery and Profile password changes can now confirm the account email with the same clear code flow.

Improved

Emergency sign-in choices stay available

Administrators can pause email codes without removing password or Google sign-in, and can keep the previous reset-link option ready as a backup.

Improved

Account screens fit more devices

Sign-in, signup, and password recovery adapt to common screen heights without making the whole page scroll.

Fixed

Mobile standards cards stay still while swiping

The five journalism-standard cards no longer rise into view on phones; the restrained desktop entrance remains.

New

Verify email with a six-digit code

Email signup sends a short-lived code and creates the account only after that code is confirmed.

Improved

Resending and code errors are clear

The signup screen shows the resend countdown, remaining attempts when available, expired-code guidance, and a clear wait message when too many codes are requested.

Improved

Google and existing sign-in stay familiar

Google users continue through Google without an extra code, while existing email users keep the same password login and recovery flow.

Daily product release

UI 1.8.15

A clearer guide to the evidence map

The landing page now explains claims, sources, evidence, quotations, and missing connections before introducing the deeper report tools.

New

Understand the Reporting Evidence Map

A new plain-language guide shows how a claim connects to its source and supporting passage, and what it means when that connection is missing.

Improved

See what each reporting signal means

Claims, sources, evidence, and quotations now have short definitions before users explore Reader, Evidence Atlas, or comparison.

Improved

One support address on legal pages

Privacy and Terms now direct questions to support@veritas-shield.org.

Fixed

Consent screens and newsletters no longer become articles

Cookie choices, access screens, and multi-topic newsletters are withheld before Reader, grading, or Atlas can combine them into one report.

Improved

Standard and Deep Scan keep the same article structure

When two extractors recover the same words, Veritas keeps the clearer human-visible paragraphs. A weak browser structure can no longer be promoted to high completeness.

Fixed

Claims and Atlas use one evidence relation set

A claim is supported only when the same saved graph contains its evidence connection, and every unsupported claim has the same explicit gap in Claims and Atlas.

Fixed

Methods stay in Reader while page furniture stays out

Reporting methods remain inspectable, while highlights, captions, reporter biographies, media promotions, funding notices, and response-request wording stay outside article claims.

Improved

Sources and quotations keep cleaner identities

Recommendation bylines, discourse fragments, generic data or footage labels, possessive duplicates, and written-statement continuations now follow stricter identity and speaker rules.

Improved

Long investigations keep more complete claim maps

The bounded claim and relation capacities are doubled, and reports still withhold a grade if an article exceeds the declared complete-graph limits.

Improved

Comparisons disclose revision and coverage boundaries

Comparing scans now warns when the same URL produced different canonical text or when either retained graph is partial.

Improved

Different articles compare without forced matches

When no claims relate across two articles, Veritas says so clearly, keeps score and dimension comparisons available, and lists every claim under the article that contains it.

Improved

Finished reports fully clear the Analyzer

After analysis, the article begins with comfortable space below the website header without leaving part of the URL controls visible.

Improved

Why it scored expands to three important claims

Why it scored now expands from one claim to the same top-three order as Claims to inspect, while keeping one focused evidence path open at a time.

Fixed

Long Atlas claims stay inside their cards

Expert List keeps long claim text inside its card and preserves the connection summary below it.

Improved

Find every evidence gap

Find gaps now lists every claim with an explicit missing-evidence connection, rather than stopping after an initial group.

Improved

Earlier focus: one clear reason first

Earlier in this day’s release, Why it scored focused on the single claim that mattered most, while Trace kept every complete source-and-evidence path. The later update above expands that explanation to three claims.

Improved

Expert Atlas keeps every claim available

Map, Chain, Timeline, and List now use the same complete claim graph, so changing the view cannot hide later claims.

Improved

Comparison keeps the same Atlas coverage

Each article keeps the complete Guided and Expert claim sets, and Compare both retains every aligned claim path.

Fixed

Story reliably opens Summary

Open report now closes any Atlas view already waiting on the page before returning to Summary.

Improved

Finished reports begin at the article

After analysis, the article identity, direct review actions, four report choices, and score remain visible with comfortable space below the website header.

Daily product release

Evidence Intelligence Engine 5.9.1 · UI 1.8.10 · Extension 1.20.5

Cleaner Reader articles and a fuller extension report

Reader leaves recommendation lists outside the article, Atlas traces scoring claims through complete evidence paths, large report numbers remain fully visible, and the extension opens Reader without a false update warning.

Fixed

Recommended stories stay outside Reader

A related-story list placed inside an article no longer becomes Reader text or report evidence, while the real article paragraphs after it remain available.

Improved

Trace every complete claim path

Atlas now lists every mapped claim that connects to both a named source and an evidence passage. Why it scored remains a shorter explanation of the result, while incomplete paths remain under Find gaps.

Fixed

Story opens the report you asked for

Open report on the final Story step now returns to Summary instead of reopening a previous Atlas view.

Fixed

Time phrases no longer appear as sources

Sentence openings such as “On Monday” and “Late Friday” remain timing context instead of becoming a person or organization in Reader, reports, or Atlas.

Fixed

Large report values remain fully visible

Score, grade, analysis coverage, and Story signal numbers now keep their complete lower and right edges instead of being clipped by their surface.

Improved

Extension results show the complete reporting picture

The popup now includes the evidence snapshot, main score drivers, all seven reporting areas, and priority claims instead of stopping at three scores.

Improved

Full report starts at Summary

Opening the full report from the extension now starts with the report Summary. Evidence Atlas remains available when the user chooses it.

Fixed

Reader opens after a completed extension analysis

A directly analyzed news page now receives the complete Reader connection before opening, so a supported extension no longer falls into a false unavailable state.

Fixed

Long extension reports scroll continuously

The signed-in extension now uses one report scroller, so one continuous gesture can reach the top or bottom without stopping at a second page boundary.

Improved

Extension sign-in uses balanced space

The sign-in window fits its account heading and form while keeping a calm margin above and below the content.

Fixed

Temporary access checks do not look like updates

The extension shows version numbers only when an update is actually required. A temporary account or connection issue now explains the real problem.

Improved

Completed reports stop at the report choices

After analysis, the four report choices settle below the website header with a small amount of breathing room.

Daily product release

Evidence Intelligence Engine 5.9.0 · UI 1.8.7 · Extension 1.20.4

Fairer scores and one connected website and extension

Verification work still earns credit without taking too much from ordinary reporting. Follow-up evidence stays connected more carefully, and the website and extension now share one light, dark, or system appearance.

Improved

Verification work has a fairer place in the score

Visible fact-checking methods still earn credit, but their absence no longer takes too much from an otherwise well-sourced article. Evidence, source clarity, and careful wording carry more of the score.

Improved

Short follow-ups keep the right evidence connection

A sentence beginning with words such as it, they, or this report can stay connected to the source immediately before it when the meaning is clear. Unrelated nearby sentences remain separate.

New

Website and extension appearance stay together

Choose System, Light, or Dark in either place and the other follows. A new extension button uses the same three choices as the website.

Improved

Completed reports settle below the title tools

After analysis finishes, Summary, Claims, Dimensions, and Full report move just below the website header, leaving the Story, Reader, and Evidence Atlas row above the visible area.

Fixed

The extension opens on the right screen

Signed-out people see sign-in first, while signed-in people see the article screen first. The wrong screen no longer appears briefly while the extension checks the account.

Improved

Analyzer recognizes a recent extension faster

Analyzer can show the last recently verified extension state immediately, then quietly confirms it again so an update, sign-out, or removal still appears quickly.

Improved

Extension pages change more calmly

Moving between the article screen and settings now uses one short, subtle transition and keeps the no-motion preference immediate.

Improved

Extension sign-in fits one screen

The sign-in page keeps the account details, form, and account link visible in the normal extension window without requiring a scroll.

Fixed

Reader leaves player leftovers outside the article

Download labels and raw video or audio player code no longer appear as article paragraphs, claims, or Evidence Atlas items.

Fixed

Atlas keeps claim text aligned

Invisible spaces copied from a webpage are normalized before analysis, so a claim stays connected to the exact Reader passage it came from.

Fixed

Complete articles keep their score

A tiny menu, media card, or webpage fragment can no longer overrule a complete article and cause its score to be withheld. Two genuinely different article bodies still stop scoring for review.

Improved

Extension sign-in is quieter

The website and extension account reminder is now one short centered line, keeping attention on signing in.

Fixed

Extension sign-in always starts at the top

Chrome no longer reopens the sign-in page partway down the form after checking the account, so the full account heading and form stay visible together.

Fixed

Extension sign-in keeps its full window

The sign-in page now gives Chrome a normal full-height layout, so it cannot collapse into a tiny empty strip below the toolbar.

Daily product release

Evidence Intelligence Engine 5.8.9 · UI 1.8.4 · Extension 1.20.0

Cleaner claims and easier report navigation

Publication timestamps no longer appear as claims, important details are easier to scan, and completed reports now open at the right place.

Fixed

Article update times stay out of claims

Standalone Published and Updated timestamps remain article information instead of appearing as unsupported claims in reports or Evidence Atlas. Real events that mention a date still remain claims.

Improved

Score explanations appear once

Summary and Full report no longer repeat the score-status message below the score. The explanation and any active limitation remain available from the information button beside Reporting-evidence score.

Improved

Important claims wait until requested

What deserves attention first starts closed in Full report and shows a simple claim count and review action instead of squeezing a long claim into one crowded row.

Improved

Full report details take less space

Evidence details and priority claims now use aligned controls and a smaller, clearer open view, so the Full report is easier to scan without losing the useful claim information.

Improved

Important claims are easier to compare

Opening What deserves attention first now shows one balanced row per claim, keeping the claim, review details, evidence connection, and Reader link close together without a large empty area.

Improved

Report tools stay within reach

Story, Reader, and Evidence Atlas are now visible beside the report title. The three-dot menu keeps only the original article, copy-link, and sharing actions.

Improved

Completed reports open at the useful controls

After an analysis finishes, the Analyzer waits for Summary, Claims, Dimensions, and Full report to appear, then gently moves them below the header so users can begin without searching down the page.

Daily product release

Evidence Intelligence Engine 5.8.8 · UI 1.8.0 · Extension 1.20.0

Safer extension updates and a calmer browser experience

Chrome Web Store updates can now move through review while the current published extension keeps working. The extension also feels more like the Veritas Shield website, with clearer screens for analyzing, reading, signing in, and changing settings.

Improved

Chrome handles public updates

Store users receive Chrome’s normal automatic updates. When Chrome knows a newer release is ready, Veritas can explain the update and take the user to the official Store listing.

Fixed

Incompatible extensions stop before analysis

If an installed version cannot collect the article details Veritas needs, Deep Scan pauses before analysis and explains the update while Standard analysis remains available.

Improved

Analysis improvements reach the website and extension together

Standard and Deep Scan use the same current extraction, evidence, claim, Atlas, and scoring rules.

Improved

The extension now feels like the website

The popup, sign-in screen, settings, progress, results, and Reader now share the website’s calm layout, larger controls, and clear groups. Decorative green lines were removed so the important actions and evidence are easier to notice.

Improved

The Store page shows the real extension

Chrome Web Store images now use the actual popup, results, settings, and Reader so people can understand the experience before installing it.

Daily product release

Evidence Intelligence Engine 5.8.8 · UI 1.7.13 · Extension 1.19.4

Cleaner articles, more accurate sources, and easier access to originals

Standard and Deep Scan now agree more often about what belongs to an article. Reports identify sources more carefully, and users can reopen or copy the original article link from History and shared reports.

Improved

Original articles stay attached to reports

Single and comparison reports now offer safe Read original article and Copy article link actions from History, live results, and shared views. Comparison reports label each article separately.

Fixed

Webpage controls stay outside the article

Cookie notices, sign-in messages, video controls, and similar webpage items are removed before Reader or the report uses the article.

Fixed

Standard and Deep Scan choose the complete article more consistently

A clearly complete article can now be accepted without unrelated text at the end making it look incomplete. Veritas still withholds the score when it cannot safely find the article boundary.

Fixed

Headings and captions do not become scored claims

Article headings remain useful Reader structure, while standalone headings, media captions, dependent clause fragments, and embedded-player descriptions stay outside the claim map.

Improved

Named sources must actually provide information

Dates, places, sentence-opening words, recipients, and long titles are less likely to be mistaken for a person or organization that supplied the information.

Improved

Documents are kept separate from people

Reports, filings, archives, dictionaries, and other records are identified as documents instead of being mistaken for people who spoke.

Improved

Follow-up quotations keep the right speaker

A quotation can now stay connected to a clearly identified nearby official or group member. Veritas leaves it unattributed when the speaker is too far away or unclear.

Fixed

Publisher credits do not replace the real author

When page information names both a reporter and the publisher, the report keeps the reporter as the author. A publisher can still be the author when the article clearly credits only that organization.

Fixed

Atlas inspection stays on the selected claim

Opening a Source, Evidence, or connected-item card now changes the inspected detail without silently returning the focused path to the first priority claim. Only an explicit claim choice, Previous, or Next changes claims.

Fixed

Job titles no longer create false sources

Words inside a person’s title or an organization’s name stay part of that identity instead of making a nearby person, party, or topic look like a source.

Improved

Opinions stay connected to the person who expressed them

Criticism and condemnation now attach to the person or organization expressing that view. Short quoted labels are not treated as full spoken quotations.

Improved

Standard and Deep Scan share the correction

Both analysis paths use the same evidence rules, so Reader annotations, claims, reports, comparison, and Evidence Atlas receive the same corrected source identities.

Improved

Difficult pages still receive the complete analysis

Ordinary articles continue normally, while pages that need extra recovery still receive the complete analysis process.

Fixed

Article links stay visible in Reader

When two ways of reading a page find the same article, Reader now keeps the linked words from the copy that still has them. Links from other parts of the webpage are not added.

Fixed

Evidence Atlas columns line up

A long claim list now stays inside the same bottom boundary as the focused path. You can scroll the claim list without making the whole Atlas uneven.

Daily product release

Evidence Intelligence Engine 5.8.6 · UI 1.7.9 · Extension 1.19.2

Cleaner articles, better evidence matching, and smoother Deep Scan updates

Reports keep article links readable, remember the Deep Scan choice without flashing, recognize older installed extensions, and clearly explain when an update is needed.

Fixed

Complete Deep Scan articles keep their score

When several methods agree on the full article, a weaker page result can no longer make that article look incomplete and remove its score.

Fixed

Story listings and access checks stay out of Reader

Collection pages and verification challenges are identified before a report is published, so a list of headlines or browser-check copy cannot be graded as one article.

Fixed

Reader and Atlas use the same saved article

Paragraphs, claims, pictures, edits, and evidence locations now stay tied to one exact article copy, so they do not shift between report views.

Fixed

Repeated introductions stay outside the report

Repeated headlines, author and date blocks, audio prompts, recommendations, and duplicate copies of the article are removed before analysis.

Improved

Long articles keep more complete claim coverage

Reports can now map up to 120 claims, while repeated versions of the same claim only get one chance to affect the score.

Fixed

Article coverage follows page format

A complete short brief can pass without being mistaken for a truncated article, while unfinished liveblogs, transcripts, galleries, and video-first pages abstain instead of receiving a premature score.

Fixed

Reader leaves out embedded media controls

Video cards, player prompts, highlight labels, and similar webpage controls stay outside the article in both Standard and Deep Scan.

Fixed

General groups are not mistaken for named sources

Vague groups, placeholders, and incomplete abbreviations are less likely to be treated as a named person or organization.

Improved

Quotation results reflect real spoken quotes

Complete quotations and clear transcript turns count, while short quoted labels and scare quotes do not distort the result.

Fixed

Atlas claims remain complete sentences

A second action stays connected to the person or thing performing it instead of appearing as a broken, unsupported claim.

Improved

Atlas distinguishes quotation support precisely

A speaker-linked quotation may support the narrow attribution that the speaker made the statement, but it never becomes proof that the quoted proposition is true.

Improved

Deep Scan is easier to choose

The Analyzer now places Deep Scan beside One article and Compare two as a compact, balanced choice with its state inside the control, an accessible switch, and restrained motion.

Fixed

Article links stay clean and clickable

Visible linked words remain safe hyperlinks in the website and extension Readers, while raw anchor tags, destinations, and attributes stay outside article prose, annotations, claims, Atlas nodes, reports, and comparisons.

Fixed

Deep Scan keeps its saved choice on reload

The Analyzer restores the saved on or off choice before the control paints, so it no longer flashes the opposite state while the page hydrates.

Fixed

Installed older extensions are recognized

Extension discovery now accepts legacy replies, prevents overlapping checks from invalidating a slower response, and treats an answering older bridge as update-required instead of missing.

Improved

Deep Scan state changes feel settled

The compact switch keeps one stable footprint while its green surface and On or Off label transition together, with an immediate reduced-motion path.

Improved

Extension updates explain the next step

An outdated extension now opens a focused update screen with installed and required versions, a clear update action, and confirmation that Standard analysis remains available.

Improved

Complete articles avoid unnecessary recovery

Complete articles move directly to analysis, while difficult pages still receive the full recovery process instead of a lower-quality shortcut.

Fixed

Topic links no longer look like source documents

A link counts as a document only when it clearly points to a study, report, record, dataset, filing, audit, or other named publication that the article actually uses as evidence.

Improved

Claims and sources stay separate

Evidence Atlas now requires a source to be something that actually supplied information or evidence. Standard and Deep Scan use the same rule.

Fixed

Pronoun-linked continuation quotes remain attributable

A complete quotation introduced or followed by a speech tag such as “she added” now resolves to the closest unambiguous nearby speaker, including a quotation that continues at the start of the next paragraph.

Improved

Automatic quote checks and Reader edits follow the same rules

Both systems use the same nearby names, source descriptions, pronouns, and speech labels. A distant word or unclear pronoun cannot create false quotation credit.

Fixed

The listener is not mistaken for the speaker

In wording such as “a source told an outlet,” Veritas identifies the source as the speaker instead of the outlet, audience, or voters.

Fixed

Quotes still connect when a listener is named

Wording such as “the governor telling voters” or “a researcher said in an interview” now works in both automatic detection and Reader review.

Improved

Follow-up quotes connect safely through words like “saying”

A complete quotation can connect to one clear nearby speaker. Short labels, interrupted sentences, and unclear speaker choices remain unattributed.

Daily product release

Evidence Intelligence Engine 5.8.0 · UI 1.7.6 · Extension 1.19.0

More complete articles, clearer sharing, and a more consistent product

Standard analysis, Deep Scan, and Reader now use the same clean article more consistently. Sharing images, search information, mobile layouts, pictures, and the extension were also brought into the same experience.

Fixed

Similar controls now move in the same way

Analyzer choices, claim filters, report layouts, and Atlas article choices now use the same calm selection animation, with a no-motion option.

Improved

Comparison Atlas clearly names each article

The choices now say Article 1, Article 2, and Compare both, and selecting one article shows its complete evidence map.

Improved

Completed analyses start with the result

One-article and two-article analyses both remove the generic Analyzer introduction after completion so the report identity becomes the first heading.

Fixed

Monthly plan amounts match the real total

Plan cards calculate the monthly equivalent from the live total and length, then clearly separate payment, access, and renewal information.

Improved

Standard and Deep Scan keep the same article structure

Both analysis methods now preserve paragraphs, headings, quotations, and lists in the same order.

Fixed

Reader restores readable paragraphs

Older reports and pages with unusual formatting now appear as readable sections without moving evidence highlights.

Fixed

One day has one changelog entry

Every change published on the same calendar day is now consolidated into one release section instead of Update 1, Update 2, and similar fragments.

Fixed

Deep Scan Reader matches its article

Deep Scan now keeps the selected text, paragraphs, links, and pictures together as one article and waits for a stable complete browser snapshot before finishing.

Fixed

Reader contains the article, not its controls

Standard and Deep Scan leave out repeated headlines, author and date blocks, share buttons, follow prompts, and player text while keeping the real article.

Improved

Standard analysis finds more modern articles

Server analysis can read more content stored in modern webpage data and verified alternate article versions without running the publisher’s code.

Improved

Extra article recovery is kept under control

Network, file size, time, workload, redirect, and identity checks limit the extra work used for incomplete pages.

Fixed

Unknown links stay unknown

When recovered article text does not include its original hyperlinks, Veritas no longer treats unrelated webpage links as article evidence or describes the missing link data as a confirmed absence.

Improved

Modern webpage data produces more complete server articles

Standard analysis can read more article text stored in webpage data, hidden article sections, and clearly linked continuations without running publisher JavaScript.

Fixed

Incomplete endings withhold the grade

A strong mid-sentence cut-off now marks the article as incomplete even below the size ceiling, while ordinary unpunctuated headings are not treated as truncation.

Fixed

Outdated extensions stop before changing a report

Deep Scan, Reader, recovery, and evidence edits now pause an outdated extension before it can create or change a report.

Improved

Repeat analyses keep one current report

Analyzing the same URL refreshes its existing History report, and existing report or comparison share links resolve that newest saved data.

Improved

Search engines now see the current product information

Page titles, descriptions, links, icons, Blog pages, and the sitemap now describe Veritas and its article reports consistently.

Fixed

The mobile first look stays three lines

The three landing-page product phrases now scale within narrow phone widths and remain one line each instead of wrapping into a longer block.

Improved

Reader pictures open with the article

Article pictures now appear by default in Standard and Deep Scan Reader reports, with a compact control available when a user prefers to hide them.

Improved

Reader keeps only useful guidance

Two redundant Reader and evidence-helper sentences were removed while accessible labels and active update status remain available.

Improved

The extension feels like the same product

The popup, sign-in, results, and in-page Reader now match the website’s design, controls, pictures, reading views, evidence navigation, and full-report access.

Improved

Every public page has its own sharing poster

Home, Analyzer, About, Methodology, Blog, guide and announcement archives, Changelog, Privacy, Terms, and shared reports now use clean 1200×630 posters with exact page messaging and page-specific Veritas artwork; individual Blog posts keep their authored covers.

Improved

Announcement covers stay varied

Announcement artwork now uses more varied styles and layouts instead of repeatedly looking like a newsroom image or a Veritas webpage.

Fixed

Page previews stay current

Link previews for Veritas pages now use the same poster as search results and shared links.

Daily product release

Evidence Intelligence Engine 5.5.3 · UI 1.7.0 · Extension 1.15.2

Clearer explanations, clearer evidence, and easier mobile use

The website now explains Veritas in simpler words. Reports more clearly separate support, attribution, missing evidence, and claims that do not appear in the other article, on both phones and computers.

Improved

The website explains the product on first look

Public pages and shared images now say what Veritas analyzes, what users receive, what they can do next, and what the score cannot prove.

Fixed

Evidence proximity no longer reads as truth support

Reports, comparison, Atlas, and exports now show same-sentence or nearby attachment separately from the typed relationship: support, contradiction, context, citation, attribution, ambiguity, or insufficiency.

Fixed

Absent claims are not evidence gaps

A claim found in only one article is labeled as absent from the other. Evidence-gap language is reserved for a claim that is present but lacks an adequate attached passage.

Improved

Comparison keeps both article identities visible

The mobile tracking rail, human claim-change labels, claim filters, and default single-report layout keep Article 1 and Article 2 oriented without implying a truth winner.

Improved

Reader preserves your place

Switching between Reading and Evidence keeps the same text in view, evidence jumps include nearby paragraph text, and phones can open evidence without scrolling to the end.

Improved

Atlas queues fit the phone viewport

Guided, Expert, and paired-article claim queues use vertical selection, compact task controls, 44-pixel close targets, and separate filters for evidence gaps and claims present in only one article.

Improved

Account and History start with the work

Compact mobile metrics and deferred editorial art keep saved reports, search, profile fields, security, and plan status above the fold while preserving the Zen Linen identity.

Fixed

Plan pricing states the real purchase

Plan cards show prepaid total prices, monthly equivalents, no automatic renewal, the access end date, and a clear way to view plans or extend active access.

Improved

Progress and setup stay relevant

Analysis names the current article, percentage, finite operation, and tab requirement; Deep Scan status uses one meaningful icon on desktop and stays hidden on mobile devices that cannot install the extension.

Daily product release

Evidence Intelligence Engine 5.5.3 · UI 1.6.3 · Extension 1.15.2

Easier-to-read motion and clearer loading messages

Landing-page movement is slower and more purposeful, while loading indicators now look and behave consistently. Reports and evidence remain still and easy to read.

Improved

The landing promise has time to be read

Four evidence-focused phrases type and pause at a measured pace while reserving their layout space; reduced-motion visitors receive the complete first promise immediately.

Improved

Marketing motion explains the reading path

A one-time evidence spotlight, scroll-linked word emphasis, and restrained numbered-feature reveals guide attention without adding another hero action or changing the calm Zen Linen surface.

Improved

Premium identity responds with restraint

The membership surface gains a small pointer-only tilt and glare with no touch or reduced-motion behavior and no motion added to account forms.

Improved

Waiting states share one language

Analysis uses a finite dot matrix, page and cover preparation use bars, and submitting actions use compact dots; every state keeps an accessible label and a reduced-motion pulse.

Daily product release

Evidence Intelligence Engine 5.5.3 · UI 1.6.2 · Extension 1.15.2

An easier Atlas, shorter explanations, and clearer account pages

Evidence paths are easier to follow, Methodology is quicker to understand, account pages are clearer, and page previews load more smoothly.

Improved

Expert Atlas is a working graph canvas

Every visible node now has fixed icon anatomy, can be dragged to reorganize the map, and keeps its connectors attached while moving; Fit restores the measured lane layout.

New

Every reading depth stays in motion

A slow, pauseable landing plane presents real Veritas report and evidence views without fabricated testimonials, decorative metrics, or motion on mobile and reduced-motion layouts.

Improved

Landing features have a calmer boundary

Why Veritas Shield keeps one full-width section line while spacing—not repeated internal dividers—separates Reader, Evidence Atlas, and Deep Scan.

New

Public pages close with an evidence skyline

A Veritas-built document city now sits as quiet background architecture beneath concise actions on a clean Linen reading field.

Improved

Landing methodology tells one story

One open Read-to-Measure-to-Explain sequence replaces duplicated methodology prompts, nested cards, and an oversized product-view showcase.

Improved

Reader signals take less space

Source, read time, length, and score now use compact inline signals without decorative status dots, leaving more room for the article and evidence inspector.

Improved

Atlas paths stay aligned

Icons, labels, details, and connecting lines now stay in place even when quotations have different lengths.

Improved

Methodology starts with the answer

A one-minute method, compact scoring view, and progressive source shelves preserve every cited link and exact rule without making first-time users cross a wall of text.

Improved

Account access feels part of the evidence desk

Log in and sign up now use an open Zen Linen composition, a quieter mode switch, and one focused form while preserving Google, email, recovery, and return-path behavior.

Improved

Analysis progress is easier to follow

The current task, progress, and three clear steps remain visible without putting every update inside another box.

Fixed

Editorial page images stay warm

Blog and announcement link previews keep the destination image ready during repeat hover, then release it when it is no longer needed.

Improved

The closing invitation has time to land

The final headline now enters at a calmer pace while keeping the same immediate, static presentation for users who prefer reduced motion.

Improved

Payment feedback is direct

Invalid references and promotion codes receive accessible field feedback, payment states slide in place, and a reduced-motion-safe success confirmation completes the workflow.

Fixed

Extension actions remain inside the dialog

The setup window scrolls when needed and keeps its download button visible on both computers and phones.

Fixed

Mobile navigation appears once

The redundant header menu is removed below desktop width, and six public and legal destinations form an even two-column footer without repeating the Analyzer action.

Improved

Mobile trial reassurance stays with the action

The three trial and evidence assurances now form one centered line immediately below the stacked landing actions.

Improved

Legal, account, History, and Blog stay consistent

Privacy and Terms return to open reading sections with their left navigation, membership status gains a restrained identity surface, History search clears in place, and editorial links share one hover language.

Improved

Account and History status now reads at a glance

Identity, sign-in methods, plan access, saved work, and comparison selection use compact signal cards with icons, status, and progress instead of plain spreadsheet cells.

Improved

Reader has one evidence workspace

The article overview and right-side evidence inspector now share the same visual hierarchy, compact evidence counts, progressive disclosures, and focused review surfaces.

Improved

Setup and payment follow visible steps

Extension installation, extension updates, plan payment, and manual review now present status first and ask for one action at a time.

Improved

Legal pages scan like the rest of Veritas

Privacy and Terms use tracked section navigation and focused reading surfaces without changing their meaning.

Fixed

One date is one changelog section

All releases published on the same day now share one dated section and one navigation target.

Improved

Profile is split by intent

Profile, Security, and Plan are now focused peer views with account status visible before any form or detail.

Improved

Reader makes evidence a clear mode

Reading and Evidence share one compact control, core article metrics stay visible, and highlight instructions wait behind a disclosure.

Improved

History scans before it explains

Report, comparison, and selection totals lead the library while comparison instructions and the final action appear only when useful.

Fixed

Compare Mode has one home

The duplicate avatar-menu switch is removed; One article and Compare two remain beside the Analyzer form where their effect is visible.

Improved

Expert Atlas starts with the graph state

A compact metric strip, focused workspace, and selected-item card replace the previous wall of controls and graph explanation.

Improved

Expert chains focus on one claim

A claim queue and one visible claim-to-object chain keep the current path readable, while additional connections remain available on demand.

Improved

Filters and exports wait until requested

Graph refinement lives in one collapsible panel, and full screen, stable links, and exports share one compact menu.

Fixed

Report tabs keep their place

Summary, Claims, Dimensions, and Full report preserve the navigation anchor when switching between panels of different heights.

Improved

Dimensions opens without redundant instructions

The dedicated Dimensions view now starts directly with the weighted scoring areas and their evidence.

Daily product release

Evidence Intelligence Engine 5.5.3 · UI 1.5.0 · Extension 1.15.2

Easier evidence paths and smoother report navigation

Evidence Atlas is easier to scan, while Summary, Claims, Dimensions, and Full report remain separate views for different levels of detail.

Improved

One article follows one visible evidence path

Atlas now pairs a short claim queue with a focused claim, source, and evidence route; complete metrics and locators stay collapsed until requested.

Improved

Two articles compare in aligned lanes

Comparison Atlas presents one paired claim path at a time, with quick filters for shared, different, and missing evidence connections.

Improved

Report views move without losing context

Summary, Claims, Dimensions, and Full report remain outside Atlas and gain a shared active indicator with restrained, reduced-motion-safe transitions.

Improved

Dimensions can be read on their own

A dedicated Dimensions view lets users scan weighted areas and open only the evidence they need without entering the complete report.

Improved

One article opens with the answer

Summary leads with the score, four evidence counts, the weighted areas with the most room to improve, and three claims to inspect.

Improved

Claims scan before they expand

The priority-ordered claim index keeps support state and evidence quality visible, then reveals the complete passage, relation, and source only when requested.

Improved

Full depth stays available

Every existing scoring dimension, Evidence Intelligence explanation, publisher detail, signal, Reader passage, Atlas route, Story, and share action remains available.

Improved

One report action leads

Inspect evidence stays primary while Story, Reader, Evidence Atlas, and sharing remain available inside one compact More menu.

Improved

Deep Scan status is quieter

Completed reports show Deep Scan as a small report detail, and the Analyzer switch no longer repeats an unnecessary Use label.

Improved

Comparison reads at a glance

The score difference, changed claims, source counts, and simple bars replace repeated explanatory paragraphs.

Improved

Claim detail opens only when needed

Paired claim previews stay short by default, while each row expands to the complete evidence, source, quality, and priority detail.

Improved

Full reports remain side by side

The synchronized full-report view and its one-article fallback remain available without competing with the default summary.

Improved

Full reports compare side by side again

The complete Article 1 and Article 2 reports render in synchronized columns by default, with focused single-report views available on narrower screens.

Fixed

Comparison context no longer covers the report

The large workspace introduction scrolls normally while a compact sticky rail keeps article identity and score aligned during detailed review.

Fixed

A failed History section no longer crashes the page

Saved report versions now load from their current evidence record, and one failed section shows a warning instead of closing all of History.

Fixed

Deep Scan status stays readable

A meaningful status icon remains beside its state label, the connection explanation receives its own line, and the action stays aligned at desktop and phone widths.

Improved

Atlas starts with three review tasks

Why the grade, trace a central claim, and find missing evidence now open a concise claim-to-evidence chain. Full graph layouts, filters, sharing, and export stay inside Expert tools.

Improved

Comparison explains the difference first

The workspace opens with the score difference, the areas that mattered most, matched claims, and their sources. Complete reports remain available for closer review.

Improved

Reports separate the grade from article completeness

Coverage describes how much of the article Veritas could analyze, not whether the article is true. The leading reason also opens at its exact Reader passage.

Improved

Analyzer modes and errors are visible in place

One-article and two-article modes sit beside the form, sign-in expectations are clear, invalid fields explain what needs attention, and one failed comparison side can be retried without discarding the successful report.

Fixed

Reader and Atlas use the full phone workspace

Reader evidence labels remain visible, previous and next controls move between passages, Atlas task controls wrap cleanly, and selected-node details stay collapsed until requested.

Improved

Story, History, and mobile reports keep the next action clear

Story shows an always-visible report exit and step context, History explains the exact two-report selection, and mobile reports surface the score before decorative artwork.

Daily product release

Evidence Intelligence Engine 5.5.3 · UI 1.4.5 · Extension 1.15.2

More complete claims and more dependable Atlas comparisons

Claims are split more naturally, duplicate pictures are handled better, sources keep the right identity, and similar claims line up more accurately across two articles.

Fixed

Captions and editor labels are not graded claims

Picture captions, update notes, developing-story notices, and short flattened promotion headlines remain inspectable context without changing claim coverage or score.

Fixed

Claim clauses remain grammatical

Coordinated objects and attribution tails no longer split into fragments, while independent assertions remain separate and multi-sentence attributed quotations keep their visible support.

Fixed

Source names keep the correct identity

People keep their titles and full names, clear surname references match consistently, and vague words such as “statement” cannot combine unrelated sources.

Fixed

One picture creates one Atlas item

Responsive transform variants and byte-identical assets collapse before graph construction, URL-declared thumbnails are rejected, and caption credits retain readable spacing.

Improved

Comparison matches the same claim across different levels of detail

Short and detailed versions of the same claim can line up, while changed numbers, units, positive or negative meaning, and direction remain clearly different.

Improved

Comparison downloads avoid repeated data

Each article is saved once in the download, and comparison rows point back to its claims and evidence instead of copying everything again.

Fixed

Details from later stories stay with those stories

Authors, published dates, updated dates, and fact-check details must match the selected article. Recommended or automatically loaded stories cannot add credit to it.

Fixed

Visible fallback evidence stays in the main story

Bylines, dates, and disclosed verification methods are searched within the selected article boundary instead of unrelated cards or surrounding page text.

Fixed

Repeated versions of one official link count once

Tracking codes and page sections no longer make the same government resource look like several sources. Different documents still remain separate.

Fixed

The Google Sign-In announcement reads correctly

The published Blog title now uses the grammatically correct “We Now Support” wording.

Fixed

Blog archive links fit small screens

Long announcement headings and archive links stack cleanly on phones instead of creating horizontal overflow.

Fixed

The same evidence cannot add points more than once

Different links or labels for the same research paper or record now count once, while genuinely different documents remain separate.

Fixed

Split extractors now abstain

A two-versus-two article-boundary tie is reported as material disagreement instead of high-consensus content.

Fixed

Headlines and dates stay with the right article

A recommended story cannot replace the selected article’s headline, and published and updated dates can be completed from the page’s available information.

Improved

Picture inspection explains what it can and cannot know

Atlas checks usable image versions and can show size, format, exact-file matches, visual matches, and available file details. Unconfirmed credentials are clearly labeled as unverified.

Fixed

Picture context stays local

A caption can contextualize only an adjacent claim with compatible visible years; distant repeated wording no longer creates a false connection.

Fixed

Claim comparison respects meaning

Percent versus percentage points, conflicting units, polarity, direction, and time scope cannot align merely because the wording and number overlap.

Improved

Mobile routes are complete

A compact site menu and the footer now expose the Analyzer and every primary public route on small screens.

Improved

Social preview is lighter

The Open Graph image keeps its original pixels while using lossless PNG recompression to reduce transfer size.

Daily product release

Evidence Intelligence Engine 5.5.0 · UI 1.4.1 · Extension 1.15.0

More faithful claims, clearer documents, and safer Reader edits

Headlines and multi-part claims keep their full meaning, research papers and pictures are described more carefully, and Reader edits now update as one complete result.

Fixed

Headline meaning cannot match on a number alone

Direction, time, and uncertainty now remain part of claim alignment, while equivalent cautious wording stays aligned.

Fixed

Each separate claim needs its own evidence

Different statements in one sentence are separated, and nearby evidence must actually support the same statement.

Improved

Journal papers become inspectable source documents

Titles, identifiers, bound links, and local method context resolve to document entities without turning a citation into proof of the claim.

New

Pictures can be inspected without changing the score

Atlas can show file matches, visual matches, dimensions, formats, and available details, but does not claim to know the picture’s origin or truth without evidence.

Fixed

Reader edits save as one complete update

Recalculation now saves the report, evidence, revision, and review history together, so the website and extension cannot see a half-finished result.

Improved

Atlas comparison distinguishes origin from independence

Evidence swaps and distinct origins remain visible without claiming independent verification that the article does not establish.

Fixed

Extraction consensus resists one outlier

Agreement among independent article candidates can select the stable article body without letting one contaminated candidate force an unnecessary abstention.

Improved

Public pages are faster and easier to read

Home no longer waits for account details before appearing, phone text stays readable, and Blog search no longer misses later results.

Improved

Compare actions keep their place

The comparison artwork now sits beside the report heading while sharing and score-difference controls remain aligned with the comparison summary.

Improved

Long-form pages show your current section

Changelog dates and Methodology sections now link directly to their content and bold the section currently in view.

Fixed

Incomplete evidence maps cannot retain a grade

Material extraction disagreement, empty claim maps, and exceeded mapping capacity now withhold the score everywhere instead of pairing a limitation with a misleading number.

Fixed

Claim details stay attached to the right assertion

Independent coordinated assertions are separated even without comma punctuation, and shared dates no longer hide conflicting quantities in headline framing.

Improved

Research references are judged more carefully

A vague mention of a study remains visible but receives less credit than a research paper with a title, link, or clear identifier.

Fixed

Different sources stay separate

Different sources named in one passage remain separate, while repeated detections of the same source are combined.

Improved

Deep Scan survives long analysis work

The extension uses the streamed analysis channel, can recover a completed saved report after an interrupted connection, and cancels work when its global deadline is reached.

Fixed

Extension Reader contains keyboard focus

Keyboard focus remains inside the open Reader and returns to the prior page control when the Reader closes.

Daily product release

Evidence Intelligence Engine 5.3.0 · UI 1.3.0 · Extension 1.14.1

A cleaner Atlas, dependable sharing, and complete score details

Promotional page text stays out of analysis, evidence uses consistent sentence locations, repeated source names are matched safely, and claims are shown more clearly.

Fixed

Related-story text stays out of the Atlas

Long all-caps link headlines and calls to action flattened into article prose are removed by publisher-independent boundary rules before claim, source, and evidence detection.

Fixed

One passage stays one evidence item

Repeated detections of the same passage are combined even when the short source label differs, while their details remain available.

Improved

Repeated source names and claim topics are combined safely

Clear surname and shortened organization references connect to the full nearby name, and repeated versions of the same claim topic share one Atlas item.

Fixed

Sentence boundaries stay aligned

Claims, quotations, sources, and locators use one sentence map, so clock abbreviations such as p.m. no longer create fragments or shift evidence links.

Fixed

Sharing recognizes the correct public report

Share links now open the correct Veritas report while rejecting missing or lookalike website addresses.

Improved

Story shows every part of the score clearly

The score breakdown shows only the areas that affect the grade, rounds values consistently, and keeps long labels from overlapping their numbers.

Fixed

Sharing recovers cleanly when a request is interrupted

Report and comparison sharing now times out clearly and offers retry and copy fallbacks instead of leaving the share action stuck.

Fixed

Claims no longer become their own evidence

Unsupported claims connect to an explicit Evidence Gap instead of a synthetic evidence item that repeats the claim text.

Fixed

Quotes and sources point to one evidence passage

A quote, its speaker, and a document found in the same passage remain visible as different details of that passage. Repeated detection cannot make one source look independent from itself.

Improved

Comparison Atlas includes both articles

Atlas keeps the full evidence map for both articles and shows each matched claim with its passages and sources. Reused sources are marked as repeated rather than independent.

Improved

Evidence quality is explained for each claim

Each claim shows whether its evidence is relevant, direct, close to the original source, open about its method, independent, and easy to inspect. Repeating one source cannot look like confirmation from several sources.

New

Claim framing gets its own score area

Veritas now checks whether the headline matches the article and whether important limits, uncertainty, and context are kept. This does not measure viewpoint or political bias.

Improved

Evidence is balanced with the rest of the grade

Evidence remains important, while framing, language, presentation, authorship, and disclosed checking practices each keep their own role in the final grade.

Fixed

Major evidence gaps now limit the grade

An unsupported headline or central claim now places a real limit on the grade instead of showing only a warning beside a high score.

New

Claim verification stays claim-level

External results use supported, contradicted, mixed, insufficient, ambiguous, stale, unsupported-domain, or unverifiable states. They remain separate from the reporting-evidence score, and Veritas does not produce a final truth percentage.

Improved

Evidence Atlas views are easier to read

Adaptive cards keep complete labels reachable, readable zoom limits preserve type size, and drag or swipe navigation moves through wider lanes. Full-screen Atlas keeps the complete workspace visible, while Map, Evidence chain, Timeline, and List each have a dedicated focus view.

Improved

Priority review becomes preview-first

The full report now previews the leading quote and priority before a View details control. Expanded cards use the same calm visual language as the Evidence snapshot.

Fixed

Story actions share one layout

Reader, Explore connections, and Full report now use matching full-width button dimensions and corner treatment on the final Story page.

New

Veritas Evidence Atlas ships now

Guided and expert modes trace claims, sources, evidence, publisher practices, and verification limits through map, chain, timeline, and accessible list views. Exact details, shareable links, downloadable data, and article comparison remain available.

New

Claim priority now includes possible decision impact

Priority considers how central, important, consequential, and checkable a claim is. Important but hard-to-check claims remain visible without being treated as true.

Improved

Every report view uses one saved evidence record

The cleaned article, claims, people, topics, sources, evidence links, publisher details, verification status, Atlas map, and Reader edits now stay together across reports, History, Compare, sharing, and the extension.

Fixed

Incomplete extraction no longer leaks a grade

Web reports, Story, Compare, the extension popup, and Reader all withhold the number and grade when the Engine abstains instead of falling back to a legacy stored score.

Fixed

Evidence cannot spill across an unrelated clause

Same-sentence position alone is no longer support. Coordinated clauses are evaluated independently, and nearby evidence still requires coherent subject-matter overlap.

Fixed

Large evidence maps stay consistent with the score

The score and Atlas use the same set of claims. When a very large article reaches a safety limit, the report says so and keeps at least one evidence connection for every included claim.

Improved

Ownership and funding stay independent

Publisher governance now stores separate ownership and funding profiles. Funder or “supported by” language can no longer become ownership evidence, while funding-model, donor, sponsorship, financial-statement, and Form 990 routes are recognized more consistently.

Improved

Public-policy detection is more precise

Decision Impact distinguishes government, public, regulatory, and domain-specific policies from routine internal company policies.

New

Verification limits are stated clearly

Veritas reports when a claim falls outside the supported verification area instead of presenting an unsupported result.

New

Reader keeps a permanent history of accepted edits

Adding, removing, restoring, or adjusting evidence records the old and new value, exact passage, reviewer, rule version, and retention information.

Improved

Public methodology matches the Engine

Methodology and search descriptions now publish the score boundary, normalized dimension weights, Claim Priority formula, graph model, abstention behavior, current Atlas availability, research basis, and verification limits.

Daily product release

Web experience · JQA 4.2.0 · Extension 1.13.11

More precise evidence and transparency checks

Sources, quotations, claim support, policies, funding, language, and Deep Scan evidence are handled more accurately.

Fixed

Repeated and generic source wording no longer inflates grades

Equivalent source occurrences now contribute one logical scoring unit, while generic roles such as analysts, companies, officials, and witnesses remain generic evidence rather than named identities.

Improved

Claims connect to evidence at clause level

When a sentence contains separate assertions, evidence attached to one clause no longer makes an unrelated clause appear supported.

Fixed

Quotation forms behave consistently

Straight and curly single-quoted speech now follows the same attribution and Reader rules as double-quoted speech without treating apostrophes as quotations.

Improved

Policy and funding disclosures are more specific

Reports recognize a broader set of structured newsroom policies and funding disclosures while distinguishing explicit corrections and funding routes from general standards or About pages.

Fixed

Deep Scan hints cannot prove themselves

Extra page details only count when they can be confirmed in the separately captured article page.

Fixed

Biography and byline evidence stay aligned

When a grounded author biography identifies the credited person or news organization, reports now preserve that identity for authorship credit while continuing to reject generic staff labels and unrelated prose.

Daily product release

Web experience · JQA 4.1.4 · Extension 1.13.9

A fuller Blog and a safer review experience

The Blog gains announcement archives, while reports, Reader privacy, account recovery, Deep Scan, and navigation become safer and easier to use.

New

Dedicated article and announcement archives

The Blog landing page keeps each stream to one recent row, while minimal archive links open searchable, paginated pages with stable URLs for the complete article or announcement library.

New

Official product and company announcements

Blog users can now browse dedicated product, version, and company updates alongside research-backed guides.

Improved

Announcement cards match the reading desk

Official updates now use the same neutral article-card surface as normal guides, without a tinted glow competing with the content.

Improved

Grading evidence stays closer to what the page shows

Article dates, documentary evidence, correction links, anonymous sourcing, and verification details follow stricter visible-evidence rules, with clearer page-level and derived explanations in Reader.

Improved

Reader is faster and more private by default

Long evidence ranges are processed more efficiently, and remote article images remain hidden until a user explicitly chooses to load them.

Fixed

Account changes require stronger verification

Sensitive password and email changes now use supported reauthentication and verified recovery steps instead of trusting an existing session alone.

Fixed

Deep Scan transfers only the article it needs

Extension 1.13.6 sends the sanitized article extraction for analysis instead of the surrounding page document.

Improved

Navigation stays centered

Desktop page links remain geometrically centered even when the account controls on the right change width.

Improved

Mobile headers stay focused

Compact headers keep only account and page controls, while one ordered mobile footer keeps About, Methodology, Blog, Changelog, Privacy, and Terms together.

Improved

Remaining access appears with plan choices

Members with an active subscription or trial see their remaining days in the plan dialog, while headers and the landing account menu stay focused on primary actions.

Improved

Public analytics collect less user data

Blog popularity uses aggregate view counts with daily browser deduplication and does not store user identity, IP address, or browsing history.

Fixed

Extension status updates without refreshing

Installing, updating, enabling, disabling, signing in, signing out, or changing Website Deep Scan listening state now updates an already-open Analyzer automatically.

Fixed

Deep Scan completes more reliably

Longer rendered-page analyses keep their connection active until the report is ready, and an interrupted browser connection can restore the exact completed report without running the analysis twice.

Fixed

More article details appear in reports

Deep Scan now retains the detected headline, byline, publication and update timing, plus visible public accountability links needed to identify and explain more articles accurately.

Improved

Public accountability details are easier to find

Reports identify correction and editorial-policy routes, ownership and funding disclosures, and genuine newsroom contacts more consistently across supported publisher layouts.

New

Remove one phrase or every match

Reader evidence review now lets you remove only the highlighted occurrence or every matching phrase when the same detected wording appears more than once.

Fixed

Blog articles paint correctly in mobile Safari

Article headers, cover images, and long-form sections now use a stable full-width layout from the first frame instead of waiting for a scroll to correct the page.

Daily product release

Web experience · JQA 4.1.4 · Extension 1.13.3

Safer extension checks, complete History, and clearer navigation

Extension account details stay private, every recent History item remains available, public pages are easier to reach, and Reader edits update reports consistently.

Fixed

Extension account details stay inside the extension

The website receives only authenticated and account-match status, while published packages no longer inject the bridge into localhost pages.

Improved

Complete History in manageable pages

Reports and comparisons load a page at a time, so a large 30-day History stays fast without hiding older items.

Improved

Public pages are easier to reach

The landing page and Analyzer now share the public navigation, including Changelog, while the footer focuses on Privacy and Terms.

Fixed

Reader edits update affected signals consistently

Partial evidence changes now preserve the remaining reviewed material and keep the visible signal explanation aligned with the recalculated grade.

Fixed

Official-record signals reject misleading links

Public-sector evidence now distinguishes authoritative records from misleading or unrelated destinations more reliably.

Fixed

Contractions remain ordinary article text

Language and quotation signals no longer misread ordinary apostrophes in contractions or possessives.

Fixed

Published fact-check details are checked more carefully

Fact-check information must be complete and believable before it appears in a report.

Fixed

Response-handling signals are more relevant

Reports now connect visible response evidence to the dispute it addresses instead of overstating generic response language.

Fixed

Deep Scan evidence stays grounded in the page

Deep Scan reports use evidence that can be verified against the analyzed article rather than unsupported extraction hints.

Improved

Report evidence is easier to operate

Critical evidence labels use larger type and interactive review controls meet a more reliable touch-target size.

Daily product release

Web experience · JQA 4.1.2 · Extension 1.13.0

Clearer evidence, safer reviews, and smoother pages

Evidence connections are more accurate, edits from one tab cannot overwrite newer work from another, and public pages feel smoother without distracting from reports.

Fixed

Claim-support signals connect evidence more accurately

Reports are more precise about whether visible evidence supports the statement it accompanies.

Fixed

Quote-attribution signals are clearer

Reports distinguish attributable quotations from unattributed quoted text more reliably.

Fixed

Reader edits cannot overwrite another tab

Evidence reviews carry a revision and update timestamp, so a stale tab stops with a clear refresh message instead of silently replacing newer annotations.

Improved

A calmer sense of depth while scrolling

Each major route gives one meaningful illustration a restrained scroll-linked lift and tilt, with a static reduced-motion experience and no moving controls or reading text.

Improved

Faster public pages and blog browsing

Public plan data is cached, blog indexes avoid full article bodies, published libraries paginate completely, and responsive image delivery reduces unnecessary bytes.

Fixed

Redirected Deep Scans stay on public articles

Extension 1.13.0 revalidates every navigation and final redirect before analysis, and no longer logs session or report payload details.

Improved

Public routes are easier to explore

Desktop and mobile navigation now expose the Analyzer, Methodology, Blog, and About routes while preserving the focused workspace.

Daily product release

Web experience

Clearer sign-in recovery

Login failures now remain visible on the page with safe next steps for people returning to their accounts.

Fixed

Login failures always explain what to do next

An invalid email or password now produces both a persistent inline alert and a notification, with safe guidance for accounts that were combined with Google.

Daily product release

Web experience · JQA 4.1.1 · Extension 1.12.3

A new Blog, safer account recovery, and safer Deep Scan

Veritas adds practical evidence guides, restores optional Google sign-in with password recovery, and checks Deep Scan destinations more carefully.

New

Evidence-first Blog

Users can browse practical media-literacy guides about sources, evidence, and context.

New

Google sign-in and One Tap

People can choose Google sign-in where available, with a clear password path for account recovery.

New

Verified Google account linking

Signed-in members can connect Google from Profile. If the Google identity belongs to another Veritas account, the verified accounts combine without losing History or share URLs, and remaining subscription time is added together.

Fixed

Reliable Google account linking

Connecting a Google identity now completes more reliably without interrupting the account experience.

Improved

A more visual reading desk

The Blog now uses the shared editorial page treatment for a calmer, more consistent reading experience.

Fixed

Accountability signals reflect applicable evidence

The accountability dimension now presents only relevant visible signals and clearer response-handling evidence.

Fixed

Safer Deep Scan targets

Extension 1.12.3 checks an article address before opening it and rejects private-network or unverifiable destinations.

Daily product release

Web experience · Extension 1.12.2

Cleaner reports without losing any evidence

Full reports, comparisons, Story, and the extension now share a calmer design while keeping every score, rule, highlight, and Reader action.

Improved

A clearer full-report reading order

Single reports in the analyzer, History, and shared links now lead from report identity to score, review priorities, dimensions, and supporting signals with fewer nested boxes.

Improved

Complete comparisons with less visual pressure

Compare Mode keeps both full reports, Stories, Readers, evidence, and recalculation controls while using a shared editorial ledger and a clearer center division.

Improved

A more balanced six-step Story

Each Story step now uses purposeful evidence artwork when space allows, and the conclusion keeps score, grade, confidence, and actions aligned on desktop and mobile.

Improved

More descriptive product illustrations

Reader, speed, methodology, report, and comparison artwork now shows the product idea it supports instead of relying on generic symbols.

Improved

One icon language across web and extension

Controls now use the same Tabler icon family for more consistent weight, shape, and meaning.

Daily product release

Web experience · JQA 4.1.0 · Extension 1.12.1

More accurate evidence and cleaner pages

The Analyzer, reports, Story, and public pages are easier to use, with more accurate checks for sources and public transparency information.

Fixed

Source-diversity signals distinguish publisher material

Reports now separate publisher-origin material from independently attributed sourcing more clearly.

Improved

Wider, uninterrupted single-article reports

For a single URL, each scoring dimension now expands to its full height with no fixed-height scrollbox, so evidence reads in one continuous flow.

Improved

Clearer analysis progress

The analyzer now shows the read, inspect, and apply steps advancing instead of appearing to stall on the first step.

Improved

More reliable transparency signals

Corrections, ownership, and contact information are represented more consistently across supported article pages.

Fixed

One action per detected passage

Selecting a passage that was detected as more than one evidence type now shows a single, correct Remove detected action.

Improved

A more coherent set of pages

The landing hero fits mobile in one view, About and Methodology steps swipe horizontally on phones, and every public page shares the same floating-artwork header, spacing, and Deep Scan status colors.

Improved

A calmer results Story

The post-analysis Story now uses the same warm, flat editorial surfaces as the rest of the product while keeping its steps, animations, and single-screen fit.

Daily product release

Web experience · Extension 1.12.0

A calmer and more consistent Veritas

The website, Analyzer, reports, account pages, and Chrome extension now look and feel like one product.

New

A unified visual language

Geist typography, warm editorial surfaces, deep-teal branding, consistent page headers, and quieter spacing now run throughout the website and extension.

Improved

Reports that are easier to scan

Scores, evidence, confidence, warnings, Reader controls, History, and shared reports now use clearer hierarchy with fewer unnecessary containers.

Improved

Purpose-built transparent artwork

New evidence illustrations sit naturally on light and dark backgrounds and explain key ideas without blocking the content.

Fixed

Consistent dialogs and messages

Download, pricing, account, warning, error, and confirmation windows now share the same readable layout while keeping the right colors for each type of message.

Improved

Claim-support signals are more precise

Reports connect factual statements with relevant visible support more consistently.

Fixed

Article evidence is separated from surrounding page links

Reports more clearly distinguish evidence inside the article from unrelated navigation or publisher-page material.

Improved

Cleaner article pictures and authors

Deep Scan and the extension more reliably reject avatars, related-story media, and appended infinite-scroll articles.

Daily product release

JQA 4.0.1 · Extension 1.10.25–1.10.37

Safer evidence edits and clearer reports

Evidence edits remain dependable after analysis updates, while reports, pictures, mobile controls, and extension update checks are easier to use.

New

Version-aware annotation history

Earlier evidence edits are revalidated after a detector update; unmatched decisions remain visible but cannot affect the current score.

Improved

Compact score and confidence summaries

The extension popup and full report make the grade, confidence, dimensions, and next action easier to scan.

Fixed

Reader pictures stay with the right paragraph

Article pictures appear beside the correct paragraph, while outdated page copies and unrelated media are rejected.

Improved

More dependable mobile and extension controls

Core actions use larger touch targets, safer website integration, clearer settings, and reliable installed-version checks.

Daily product release

Extension 1.10.21–1.10.24

Precise evidence edits that can be undone

Reader evidence can be removed or restored more exactly without losing the rest of the review.

New

Partial or whole-signal removal

Users can remove a whole detection or a precise passage while the remaining evidence stays active.

Improved

Grouped evidence decisions

Active and removed text stays together as one evidence item with clear remove, restore, and recalculation actions.

Fixed

Reading position is preserved

Recalculating an evidence edit no longer moves the user away from the passage under review.

Daily product release

Extension 1.10.14–1.10.20

Article pictures, captions, and safe sharing

Reader shows more of the original article, and completed reports can be shared without giving anyone access to the owner’s account.

New

Reader pictures and captions

Web and extension Readers preserve lead and in-article images while filtering author avatars and unrelated media.

New

Thirty-day analysis sharing

Reports and comparisons can be shared as read-only links that expire with the original History item.

Improved

Faster evidence view

Evidence matches are cached and dense grading details load only when opened.

Improved

Clearer public and account pages

Public explanations, sign-in, recovery, and release notes use a more consistent hierarchy and more precise promises.

Daily product release

Extension 1.10.12–1.10.13

Clearer analysis progress and a more reliable Reader

The six analysis steps are easier to follow, and both Readers stay focused while keeping full score details available.

Improved

Calmer analysis progress

Analysis keeps its six-step Story with clearer progress, less motion, and direct paths to the report and Reader.

Fixed

Reliable multiline annotations

Highlights remain visible across wrapped lines, sentences, and source line breaks.

New

Delete and recalculate

Reader annotations can be deleted explicitly, then the report, History, and Compare results update together.

Fixed

Matching Deep Scan articles

Direct extension scans and website Deep Scan keep the same clean article as the scoring and Reader source.

Daily product release

Extension 1.10.6–1.10.11

One Reader experience, clearer prices, and better repeat analysis

The website and extension now offer a more similar reading experience, while plan prices and repeated article analysis behave more predictably.

Improved

One Reader experience

The website Reader now follows the extension Reader’s focused article layout, evidence controls, text sizing, and current score information.

Fixed

Newest repeat analysis

Reanalyzing an article refreshes its existing 30-day item and moves it to the top without creating a duplicate.

New

Promotion codes and live pricing

Current plan prices and eligible discounts remain visible from selection through payment review.

Improved

Clearer quotation signals

Attributed quotation findings now appear more consistently in the report’s signal and dimension explanations.

Improved

Better accessibility and search support

Page information for search engines, keyboard focus, skip links, search access rules, and screen-size behavior now follow one consistent standard.

Daily product release

Extension 1.8.1–1.10.5

Private History, article comparison, and Reader review tools

Reports, cleaned articles, evidence review, and the public explanation of scoring now work together in one connected workspace.

New

Private 30-day Analysis History

Signed-in web and extension reports save to one searchable private library and expire automatically.

New

Side-by-side comparison

Two saved reports can be compared across scores, Stories, Readers, dimensions, signals, and evidence.

New

Interactive evidence annotations

Users can select missed quotes, named sources, documents, or wire credits and see validated recalculation across the report.

New

Public scoring explanation

Users can now inspect what each signal means, how much each score area counts, how to read the result, and what automated scoring cannot prove.

Improved

Cleaner, more complete articles

Analysis compares multiple article candidates, removes more page noise, preserves long stories, and separates article links from surrounding navigation.

Daily product release

Extension 1.7.0–1.8.0

More complete Deep Scan articles with clear warnings

Deep Scan now compares several possible article versions and tells users when it is not sure that the extraction is complete.

New

Better article selection

Deep Scan compares multiple readable versions of the page before choosing the most complete article.

Improved

Complete clean article

Long stories are preserved and links inside the article stay separate from broader page evidence.

New

Extraction confidence

The selected method, confidence, completeness, and warnings are surfaced, while access-challenge pages are rejected instead of graded.

Daily product release

Web 3.4.0 · Extension 1.6.5–1.6.8

Clearer scores and safer extension analysis

Scores focus on evidence that users can inspect, while browser analysis, account recovery, and payments receive stronger safeguards.

Improved

Evidence-first scoring

Transparency and sourcing dimensions now explain the visible article-level signals contributing to their weighted results.

Improved

Safer extension workflow

Extension requests are matched to the correct action and limited safely, downloads are checked, and article-page support loads only when needed.

New

Account and payment recovery

Password recovery, traceable payment requests, calendar-based subscriptions, and consistent plan pricing were added.

Release archive

Web 2.15.0 · Extension 1.6.x

Clearer extension updates

People can now see what changed before downloading or updating the extension.

New

Release notes before download

Download dialogs and upgrade prompts show the published version and its user-visible changes.

Improved

Extension upgrade flow

The Tool page detects older downloaded or installed versions and presents a clearer upgrade notice.

Fixed

Clearer account-session requirements

Email changes and extension reauthentication now explain more clearly when a fresh session is required.

Release archive

Web 2.14.0

A new visual style and stronger browser analysis

Veritas received a new look and a more capable extension mode for difficult article pages.

New

Liquid Glass design

Frosted surfaces, glowing borders, dark mode, and motion created the previous visual identity.

New

Advanced Extension Mode

Article and author discovery improved coverage for complex page layouts.

Improved

Context-aware analysis guidance

Warnings explain the limitations and strengths of each analysis mode.

Release archive

Web 2.13.0

Accounts, free trials, and plans

Veritas added personal accounts, a free trial, and paid access plans.

New

Authentication and subscriptions

User accounts, 15-day trials, and flexible subscription access were introduced.

Improved

Modern responsive interface

The website gained a more consistent responsive layout and improved dark mode.

Release archive

Web 2.12.0

Chrome extension launched

People can now analyze an article directly from the page already open in Chrome.

New

Chrome extension

The first browser extension supported analysis of articles available in the current tab, including many paywalled pages.

Improved

Faster analysis

Rule evaluation was optimized for noticeably quicker credibility reports.

Release archive

Web 2.11.0

More reliable article links and phone layouts

Article links are checked more clearly, and the website works better across different screen sizes.

Fixed

URL validation

Article links receive clearer parsing, validation, and error feedback.

Improved

Responsive design

Touch interactions and small-screen layouts were improved.

Release archive

Web 2.10.0

The first detailed article score

The first recorded release introduced a fast article report with a score users could inspect and understand by category.

New

Multi-dimensional scoring

Reports evaluate sourcing, authorship, accountability, language, genre clarity, and publication context.

New

Fast article analysis

Credibility scores arrive with a detailed breakdown rather than a single unexplained number.

You’ve reached the beginning of the recorded product timeline.

Analyze your next article

Paste a news link to see its major claims, visible sources, evidence gaps, and explainable reporting score in one workspace.

Analyze an Article FreePreview the Analyzer