LoginSign Up
← Back to EdTech
Weekly · EdTech · Week of October 12, 2026

Code.org kept shipping near-daily AI Lab releases while Kolibri and TAO Testing sat in alpha and LTS mode.

ai-literacy-toolingstructured-courseslts-hardeningexplainable-models
O
Generated 2h agoDrawn from 5 products

The week in lms-edtech

Five products updated, but only Code.org shipped at real volume. Its work is concentrated on a single surface, AI Lab, where students train, test and now trace the path of a model's prediction.

The rest of the sector moved on release-cycle rhythm: alphas at Kolibri, patches at TAO Testing, and a webinar listing at D2L Brightspace that carries no release detail.

Leaders

Code.org published twelve improvements, nearly all to AI Lab inside Lab2. This week's AI Lab releases traced decision-tree paths, added a results scorecard and a floating Test panel, and moved Adaptive to unified model calls. The through-line is explanation: students are being given the means to say why a model predicted what it did.

Scribe sits at the edge of this sector as a process-documentation tool, and its spark is Autocapture, which watches work across approved business apps and turns it into reviewable guides without anyone pressing start. It feeds the Optimize agent, whose workflows teams can now split and restructure.

Kolibri put out its first 0.20 alphas, introducing assigned courses with a welcome page, units, lessons and pre/post tests. They are in QA, so this is a preview of structure the offline platform has not had before.

Wildcards

TAO Testing is in hardening mode: backports, ACL migration fixes and a manifest tagged for the 2026-11 LTS. Its one feature, scoped directory search with a page-size cap, points to installs with very large user directories.

Themes that compounded

  • Explanation of how a system reached its output at Code.org and, in a business setting, at Scribe.
  • Sequenced, assessed paths replacing flat libraries in Kolibri.
  • Scale and stability work for large institutional users at TAO Testing.

Watch this week

Watch whether Kolibri moves from 0.20 alpha 1 to a beta, which would settle whether the course flow holds up in QA. TAO Testing should keep patching ahead of the November LTS tag. D2L Brightspace is not a lead: its feed has no entries describing individual features.