Jump to
- June 24
- Quality Improvements
- Bug Fixes
- June 17
- Feature Releases
- Quality Improvements
- June 11
- Feature Releases
- Bug Fixes
- June 3
- Quality Improvements
- Bug Fixes
June 24th, 2026
📣 From the CPO
Hear from Sarah Phoenix as she explores the latest release and what it means for our customers.
This week's release expands Conversational AI Learning (currently in beta) across more of your content library while delivering a meaningful set of reliability and administrative improvements, from data accuracy in reporting to expanded global commerce support. To enable Conversational AI Learning or any of our beta features, navigate to Apps > Labs.
Read More Here ↓
Conversational AI Learning (Beta) Reaches More of Your Content
For customers participating in the Conversational AI Learning beta, we’re expanding the range of content the AI can index and surface in learner responses. With this week’s release, the AI now covers discussion boards, notebooks, list rolls, and flip card sets, in addition to content types indexed in prior releases. In coming weeks, we’ll also add indexing for workbooks, meeting information, and assets stored in your Central Library. As a result, learners asking questions through Conversational AI Learning can draw on a much broader slice of your content catalog, including reusable library assets that weren't previously part of the indexed corpus.
Beta participants also now have access to a new AI Usage Dashboard at /learn/manager/ai-usage, giving admins visibility into query volume, daily usage trends, and a downloadable query history, with no limits, charges, or billing UI during the beta period.
Better Control Over Search Engine Visibility
Unreleased courses, learning paths, bundles, and custom pages now automatically receive a noindex directive, ensuring draft content isn't indexed by search engines before it's intended to go live. The platform's XML sitemap also now filters to published content only. Both changes apply automatically, no configuration required.
More Reliable Operations and Reporting
This release addresses a meaningful cluster of data accuracy and workflow gaps that have surfaced through customer feedback. Reporting is more trustworthy, common admin tasks require fewer workarounds, and the learner experience is more consistent, with additional currency support rounding out global commerce capabilities for eCommerce 2.0 customers.
We released the following quality improvements:
Expanded Content Indexing for Conversational AI Learning
Conversational AI Learning now indexes additional native authoring content page types, making more of your platform's content available through AI-assisted search and recommendations. Newly supported types include discussion boards, notebooks, list rolls, and flip card sets. Several more content types will be added over the coming weeks, including Asset Library items. This update applies automatically; no configuration is needed.
Noindex Added to Unreleased and Unpublished Content
Search engines will no longer index unreleased or unpublished content, such as draft courses, learning paths, and custom pages. Previously, these pages were crawlable by default, potentially exposing content to the public before its intended release.
This update automatically applies a noindex directive to unreleased courses, learning paths, bundles, and pages, ensuring your content remains private during development without the need for manual workarounds. Released and published content remains unaffected and will continue to be indexed normally.
AI Usage Dashboard
Managers in the Conversational AI Learning Open Beta can now access an AI Usage Dashboard at /learn/manager/ai-usage. The dashboard shows query counts, daily usage trends, and supports CSV export — providing visibility into AI activity ahead of GA billing. No limits, payment prompts, or charges apply during the beta period.
Exclude Unpublished Content from Sitemaps
We have resolved an issue where URLs for unreleased or unpublished content were incorrectly included in the platform's XML sitemap. The /sitemap.xml output now strictly filters for published content, ensuring search engines only crawl active pages and preventing conflicting indexing signals.
Require Terms & Conditions acceptance in the Forgot Password flow (opt-in)
Today, Thought Industries registration flow requires learners to accept Terms & Conditions on the /register page before setting their password, but users who go directly to Forgot Password can reset their password without going through that same step. For organizations with requirements around Terms & Conditions acceptance, a new opt-in setting extends that enforcement to the Forgot Password flow
Admins can now close this gap with a new per-company setting: Require Terms Acceptance on Password Reset, available under Settings > Setup > Primary.
When enabled:
- Learners will be asked to confirm their acceptance of T&C the next time they reset their password. Once confirmed, their acceptance is recorded and they will not be prompted again on future resets.
- New learners created going forward will have their T&C acceptance recorded at registration and will not be prompted during password resets.
This setting is off by default. Organizations that don't enable it will see no change to their current password reset flow.
Improved Reporting Reliability: Historical Timestamp Backfill
We are completing a data quality improvement to ensure all records in our platform and reporting systems have valid Created Date and Updated Date values. These fields drive key platform functionality — including status group rules — and are foundational to accurate reporting. Records created during data migrations, such as users or content imported during initial onboarding, could have null values for these fields because the original timestamps were not always present in the source data at the time of import. We have since implemented safeguards to prevent this going forward; this backfill closes the historical gap.
The backfill is currently in progress across both our application database and reporting database (BI Connector / Looker) and is expected to complete within the next couple of working days. Where the Created Date is available, it is used to backfill Updated Date. Otherwise, Created and Updated are backfilled with January 1, 1970 — a universally recognized placeholder date in data engineering that signals the original timestamp was not captured. This is intentional and follows industry-standard practice for sentinel date values.
Want to learn more? Head on over to developer.thoughtindustries.com for technical resources covering platform data, reporting, and integrations.
Note
Admins may see a Created Date of January 1, 1970 on certain historical records in BI Connector and Looker — most commonly on user records and content imported during onboarding. This date does not appear anywhere in the learner-facing interface. We recommend treating it as “creation date unknown” in any reporting logic that filters on Created Date.
Events tab in Learner Access widget now sorted by date
Great job, Community Engagers! This idea rose to the top thanks to your votes and upvotes, and now it's here!
The Events tab in the Learner Access widget now displays upcoming events in chronological order, followed by past events in reverse chronological order. This ensures the next upcoming event is always at the top when a learner opens the tab.
We released fixes for the following:
Learning Path Milestone Reporting Inconsistencies After Configuration Changes
We’ve resolved an issue where editing a learning path milestone, such as switching between session-based and course group-based completion, or removing courses from a milestone — could cause downstream reporting to reflect outdated or incorrect data. For example, reports might continue to show courses that had since been removed from the milestone. The root cause was that milestone IDs were sometimes regenerated when the milestone’s composition changed, breaking the link that allows updates to cascade through reporting. Milestone IDs are now stable across configuration changes, ensuring that edits to learning path milestones are accurately reflected in reporting going forward.
In the coming days, we’ll also be running a backfill to correct any historical reporting data affected by this issue.
Reliable Custom Page Slug Reuse
We have resolved an issue where reassigned page slugs would incorrectly redirect to archived content instead of the new page. The platform now automatically invalidates stale redirects and clears the associated cache during page creation, ensuring URLs always resolve to the active owner of the slug.
Security: Authentication Tokens
To reduce cross-site scripting (XSS) exposure, the signed authentication token is no longer available to client-side JavaScript via window.CONF.authToken. Authentication now travels in a secure, HttpOnly session cookie (SameSite=Strict, Secure) that scripts cannot read.
What this means for you:
- You may be signed out once. Existing sessions are invalidated by this change, so users may need to sign in again after the update.
- Custom code must stop reading
window.CONF.authToken. Any custom scripts or integrations that used it (for API/GraphQL calls, real-time updates, SCORM, or page tracking) should instead rely on the session cookie, which the browser sends automatically on same-origin requests. The token is no longer exposed to the page. -
window.CONF.cspNonceis unchanged and remains available.
Quiz and Test Preview Error
We resolved an issue where previewing a Quiz or Test page from the "Use Existing Page" tab in the Course Builder triggered a "Processing Error" alert. Previews now open immediately without error, ensuring a smoother workflow when reusing assessment content.
Quiz Review Label Correction
We resolved a visual issue where correct answers in quiz reviews incorrectly displayed duplicate labels (e.g., "Correct Answer Correct âś“"). This fix ensures that only a single, accurate "Correct Answer" label is shown when learners review quiz results after exhausting their attempts, providing a cleaner and more professional experience.
Improved Accuracy for Course Seat Tracking
We have resolved an issue where the Seats Used metric in content-level reports could drift downward and display inaccurate totals. This was caused by a processing error that incorrectly decremented the seat count when non-student roles or disabled users had their access revoked.
The platform now performs a filtered recount during both enrollment and revocation to ensure seat totals precisely reflect the number of active students.
Support for Multiple Banner Ad Widgets
We resolved an issue where adding multiple native Banner Ad widgets to a single layout would result in only one advertisement rendering. All Banner Ad widgets on a layout now initialize independently, allowing administrators to display multiple unique ad units on a single page.
Meeting Dropdown Sync for Multi-Session ILTs
We have resolved an issue where the meeting dropdown would display stale data when learners switched between different sessions of the same pre/post ILT content item. The meeting page and inline info now correctly refresh to show the specific meeting details for the active session, ensuring learners join the correct event.
Reporting Hub Pin Indicator
Fixed an issue in the Reporting Hub where the pin icon did not visually change when pinning a dashboard or look. The icon now correctly toggles color to provide visual confirmation when an item is pinned or unpinned.
June 17th, 2026
📣 From the CPO
This week, Sarah Phoenix shares how the latest release fits into the bigger picture of where our product is headed.
This release introduces a major overhaul of our technical documentation to better support your development teams, alongside automated SEO enhancements that ensure your catalog remains accurate and visible to search engines without manual effort.
Also note the final reminder about our upcoming scheduled maintenance period scheduled for this weekend.
Read More Here ↓
Comprehensive Technical Documentation Update
We have completely refreshed our technical documentation to provide clearer, more actionable guidance for developers and admins. This update ensures that your team has the most accurate information for managing integrations, customizing the platform, and scaling your learning business effectively.
Automated SEO & Catalog Visibility
Keeping your catalog visible in search engines just got effortless. Sitemap V3 (US only — EU coming soon) and automatic canonical URLs mean your courses, bundles, and learning paths are always indexed, always current, and always getting credit from search engines — with zero manual work required on your end.
Note
Reminder: Scheduled Platform Maintenance — Saturday, June 20
Â
As a reminder, we have scheduled platform maintenance this upcoming Saturday, June 20 to upgrade our SCORM infrastructure. You can expect two regional cutover windows back-to-back, beginning with our EU region at 3pm Pacific time, followed by an evening window for our US region — and your account will only experience the window for your own region.
This proactive update ensures continued stability and performance for your learning content.
You can review details about the updates and maintenance window in the June 3 release notes below, and on Saturday you can monitor live updates on our Status Page.
Developer Tools
New: Redesigned Thought Industries Developer Hub
We've launched a completely redesigned Thought Industries Developer Hub at developer.thoughtindustries.com to help your technical teams build, integrate, and scale faster. The new hub brings every developer resource together under one intuitive navigation structure, with new tools and guides designed to make complex integrations and customizations easier to implement.
What's new:
- Expanded Guides & Tutorials — New end-to-end, step-by-step walkthroughs for common integration workflows, including Creating and Updating Courses via API, SSO Integration, Webhook Configuration, Bulk Operations, and External Ecommerce Integration — so your team can go from concept to working integration faster.
- Personalize-as-you-go API examples — A new API Credentials bar lets you drop in your subdomain and API key once and see the code examples on REST endpoint pages update to match your instance, making it faster to test and adapt calls to your environment.
- Copyable, multi-language code snippets — Tabbed code samples (such as GraphQL and cURL) with one-click copy lets you grab the exact snippet you need in the language you're working in.
- Unified information architecture — Helium, GraphQL, REST, BI Connector, Integrations, and Guides now live in a single, consistently organized sidebar, so resources that used to be scattered are now easy to find in one place.
- Site-wide documentation search — Search across all technical documentation from anywhere in the hub to jump straight to the page you need.
- Discover with AI — A new AI-powered assistant lets you highlight anything on a page, ask a question in plain language, and get a grounded answer with citations back to the docs to help you explore and build faster.
- Modernized platform — Rebuilt on a modern documentation stack (Docusaurus 3 / React 18) for faster page loads and a smoother, more reliable browsing experience.
And this is just the beginning. We're already building on this foundation — a reusable CSS component library, a JavaScript library, and more developer tooling are on the way, with much more to come.
Explore everything at developer.thoughtindustries.com.
Quality Improvements
We released the following quality improvements:
đź§ Sitemap V3
Your course catalog now shows up better in search engines. Sitemap V3 (US only — EU coming soon) handles the technical setup automatically, so learners can find your content without you having to do anything.
- Always-current sitemap — Every released course, bundle, learning path, and custom page is automatically included in a search-engine-ready sitemap that refreshes within minutes of any catalog change. Search engines discover your content faster and more reliably — no manual submissions, no stale indexes. Unreleased content is never included./li
- Automatic canonical URLs — Every page now gets a canonical URL injected automatically. Canonical URLs tell search engines which version of a page is the authoritative one, preventing duplicate-content penalties and ensuring all ranking credit flows to the right page.
June 11th, 2026
📣 From the CPO
Chief Product Officer Sarah Phoenix puts this week’s release in perspective, explaining how it contributes to our evolving product strategy.
This week’s release introduces a transformative way for administrators to manage their academies through our new Claude Marketplace integration, while continuing to advance the learner experience with agentic AI capabilities and modernized platform security.
Read More Here ↓
Accelerated Content Operations via Claude Marketplace
The new Thought Industries for Claude integration significantly reduces the time and effort required to manage academy content. By enabling administrators to perform complex authoring tasks—such as drafting courses and updating lessons, through natural conversation, we are removing the traditional UI friction that can slow down content production. This allows teams to move from idea to draft in seconds, while maintaining full administrative control through a "draft-first" workflow that ensures no content goes live without explicit approval
Agentic Capabilities in Conversational AI Learning
We have introduced the first agentic function within our Conversational AI Learning feature, allowing the AI to move beyond simple conversation and take direct action on behalf of the learner. Through new interactive action cards, the AI can now facilitate and confirm course enrollments entirely inline. Successful enrollments trigger a visual confirmation with a direct link to start learning. This marks a significant shift toward a more autonomous and helpful learner experience.
Note
Reminder: Scheduled Platform Maintenance — Saturday, June 20: As a reminder, we have scheduled platform maintenance on Saturday, June 20 to upgrade our SCORM infrastructure. This proactive update ensures continued stability and performance for your learning content. You can monitor live updates on our Status Page.
OPEN BETA
Thought Industries for Claude – Early Access 🤖
As part of our investment in agentic authoring experiences, we're introducing Thought Industries for Claude in early access. Teams can now interact with academy content and authoring workflows directly from Claude using a plugin.
- Author and manage academy content directly in Claude — browse your catalog, create course drafts, and rewrite lesson content through natural conversation, no complex UI navigation required.
- AI-assisted, draft-only writes — every update stays in draft until you manually publish in the TI admin UI, so nothing goes live without your explicit approval.
- Three tiers of intelligence built in — Actions for quick catalog lookups, Routines for structured workflows like outline generation and lesson rewrites, and Playbooks for full end-to-end authoring sessions.
To give this a try, check out this article.
OPEN BETA
Enrollment Action Cards in Conversational AI LearningÂ
We introduced interactive action cards in the Conversational AI Learning experience to provide clear, inline confirmation for nudges and course enrollments.
Successful enrollments show a confirmation card with the course name and a direct link to start learning. If enrollment fails (e.g., full sessions, unmet prerequisites, restricted access), learners will see a dedicated error card with actionable next steps like joining a waitlist or viewing required prerequisites.
Read more about Conversational AI Learning here.
Bug Fixes
We released fixes for the following:
Basic eCommerce Account Page Access After Subscription Cancellation
We resolved an issue where learners on Basic eCommerce (Stripe-direct) instances were unable to access their account page after canceling a subscription. Previously, the account page at /learn/account would fail to load or hang on a loading spinner following a cancellation, preventing users from managing their profile, payments, or other subscriptions.
With this fix, the account page now renders correctly post-cancellation, and canceled subscriptions accurately display their "active until" date. We also implemented defensive guards to ensure the user interface remains accessible even if individual data points are missing or encounter errors during loading.
Improved Reliability for Bulk Event Imports
We have resolved an issue where virtual and in-person events created via the bulk import tool sometimes failed to appear in the Events List search results. Previously, these events were created in the database but remained "invisible" in the UI until a manual search reindex was performed.
With this fix, imported events are automatically indexed and become searchable immediately after the import job completes. Additionally, we have updated the bulk import tool to ensure that any background processing errors are accurately reported as "Failed" in the job logs, providing admins with better visibility into the success of their imports.
Improved Survey Reporting Accuracy After Content Modifications
We have resolved a reporting issue where learners who had already completed a survey would trigger a duplicate, blank entry in reporting if they revisited the survey after an administrator had modified the questions.
This fix introduces server-side idempotency to ensure that the platform correctly identifies and presents a learner’s existing completed attempt rather than creating a new, empty record upon page load. This ensures that survey data remains accurate and free of misleading "null" entries in both the platform UI and administrative reports (Redshift).
Consistent Quiz Review Styling for Multi-Select Questions
We resolved a visual inconsistency in the quiz review state where correctly answered multi-select questions (Select All That Apply) were incorrectly outlined in red. While learners correctly received credit for these answers, the post-quiz review now properly displays green success styling to match the active attempt state, eliminating confusion for learners and Admins.
June 3rd, 2026
📣 From the CPO
This week, Chief Product Officer Sarah Phoenix highlights the ideas and priorities that shaped the latest release.
You’ve told us that asset manager usability and availability across the platform is a multiplier for your content management efforts. This week, we’re continuing our investment in this area by streamlining the content creation workflow with a modernized and more intuitive asset management experience.
📆 Scheduled Platform Maintenance — Upcoming Saturday Windows
We're planning two Saturday maintenance events over the coming weeks to upgrade our SCORM infrastructure — the engine that powers all SCORM and xAPI content playback, and its underlying database. These upgrades improve performance, security, and long-term reliability.
Important
Important: Phase 1 below involves platform-wide scheduled downtime, not just SCORM/xAPI — please plan around the full Phase 1 window for your region.
Read More Here ↓
To minimize risk and reduce the number of Saturday events impacting your learners, we're delivering this work in two separate Saturday events with a stabilization period in between:
Phase 1 — Scheduled Downtime for Combined SCORM Engine + Database Upgrade
| Category | Details |
| Date & Time |
Jun 20, 2026 (Saturday) Two regional cutover windows back-to-back, beginning with our EU region at 3pm Pacific time, followed by an evening window for our US region — and your account will only experience the window for your own region. |
| Outcome | Both the SCORM engine upgrade (security patches, performance gains, and better monitoring capabilities) and the underlying database modernization, delivered on the same Saturday. We are utilizing the same reliable in-place methodology successfully employed for our Apr 1, 2024 application database upgrade. |
🔎 What to Expect
- During Phase 1's regional cutover, the entire Thought Industries platform may be temporarily unavailable — including authoring tools, learner dashboards, registration, checkout, reporting, and integrations, in addition to SCORM and xAPI playback. We recommend treating the full 3-hour regional window as a scheduled platform maintenance period.
- Comprehensive data integrity is guaranteed for all student scores, progress, completions, and bookmarks across both events.
- Active sessions during Phase 1 (SCORM/xAPI courses, registrations, checkout flows, API integrations) should be retried after the maintenance window. SCORM and xAPI sessions resume from the learner's last bookmark — no progress is lost.
Phase 2 — Further SCORM Performance Optimization
| Category | Details |
| Date & Time | Schedule TBD, no sooner than the Saturday after Phase 1 stabilizes |
| Outcome | A much smaller event focused on faster SCORM engine startup and automatic scaling during traffic spikes. This change will not cause or require downtime. |
đź“‹ Action items for your team
- Generally, no action is necessary. Content for SCORM 1.2, SCORM 2004, and xAPI will remain fully operational after the upgrade — no changes to uploading, launching, scoring, or reporting.
- Please plan around Phase 1's regional 3-hour window as full-platform maintenance (not SCORM-only).
- To help us manage the schedule, please reach out to Success at success@thoughtindustries.com if you have major SCORM-related events or critical platform usage planned for any upcoming Saturdays.
Note
If you have any questions, please reach out to Success at success@thoughtindustries.com or your Account Manager.
Quality Improvements
We released the following quality improvements:
New Asset Library Panel Experience
We’ve redesigned the Asset Library side panel, the tool used to insert assets directly into course content, to provide a cleaner, more intuitive experience for course builders. This update focuses on improving scannability and reducing the clicks required to find and manage your assets.
Key enhancements include:
- Streamlined Asset Table: We’ve optimized the layout by removing redundant columns and allowing tags to wrap naturally under asset titles. This ensures your most important information is always readable without truncation.
- Instant Asset Previews: You can now hover over any row to see a type-aware preview of the content (such as image thumbnails or text snippets) before inserting it, ensuring you have the right file every time.
- Direct Actions: We’ve replaced nested dropdown menus with one-click "Insert" and "Edit" buttons, speeding up your content creation workflow.
- Improved Feedback: A new confirmation toast appears immediately after an asset is added, providing clear visual feedback that your action was successful.
Note
These changes apply specifically to the asset selection panel within the course authoring experience and do not change the main Central Asset Library management page. This feature is available for users with Central Assets permissions enabled.
Bug Fixes
We released fixes for the following:
Corrected Redirects for Panorama Custom Domains
We have resolved an issue where learners using Panorama instances with custom domains were intermittently redirected to incorrect sign-in or sign-out pages. This occurred primarily when a user had a previous session cookie, causing the platform to incorrectly prioritize the parent site's host configuration over the specific Panorama domain.
With this fix, the sign-in and sign-out flows now consistently respect the custom host of the Panorama, ensuring learners and admins are directed to the correct authentication pages and maintaining a seamless login experience across all domains.
Reliable Course Completion Webhooks for External Activity
We have resolved an issue where course completions marked
via
the
POST /v2/courseExternalActivity API endpoint
did
not consistently
trigger the expected webhooks.
Previously, certain parameter combinations, such as providing
transcript
fields or certificate URLs, would issue a certificate
but fail
to update
the course status to "completed," preventing the
courseCompletion webhook from firing. This
fix ensures
that
all successful API-driven completions accurately update
learner
status
and
trigger both courseCompletion and
courseAction
events. Additionally, invalid certificate requests now
return
clear error
messages instead of failing silently, improving troubleshooting
for integrated
workflows.