Skip to main content

The news desk

AI news, checked against the source.

What changed in AI, one story at a time. Each one links to its primary source, and facts stay apart from interpretation.

Last updated ·

Thursday, November 19UTC

  1. New
    MicrosoftAion Instruct in EdgeModels

    Microsoft puts its small Aion model in Edge and opens a web trial

    Microsoft made its 2.3-billion-parameter Aion Instruct model generally available as Edge’s built-in small language model, replacing Phi-4-mini, and released model weights under the MIT license. Web developers can test an on-device Prompt API through an origin trial; a separate Decisions API remains an experimental preview. Edge downloads the model only when needed on supported devices, and users can disable on-device AI. Windows integration remains planned rather than available.

    Microsoft2 sources

Free, email only

Unlock the full brief free.

  • Thirty days of stories to browse, not seven
  • On every story: what to check, each source and why it matters
  • The same signup unlocks every free tool

Continuing subscribes the email on that account to Work Worth Doing, my newsletter, the same as the form below, and unlocks every free tool in this browser. We only receive your email address. Unsubscribe in one click.

This is not an account: there is no password, and the unlock is a cookie in this browser. You also join Work Worth Doing, Demetri Panici's newsletter, every week: what I built and what changed in AI. We'll email you a link to confirm, and you can unsubscribe in one click. The same signup unlocks every free tool on the site. How your email is handled.

Already subscribed? Enter the same email to unlock this browser. You won't be signed up twice.

Tuesday, November 17UTC

  1. New
    MicrosoftMAI Image 2.7 / FlashImages

    Microsoft puts MAI Image 2.7 and Flash in public preview for image editing

    Microsoft made its MAI Image 2.7 and Flash models available in public preview on Foundry and in MAI Labs, including a new Canvas editor. The company lists about $35 per 1,000 images for 2.7 and $10 for Flash in Foundry. Microsoft’s quality rankings and partner usage are its claims; the models are made by Microsoft even when offered through other apps.

    Microsoft AI1 source

  2. New
    MicrosoftCode 2 and Code 2 FlashModels

    Microsoft launches its own Code 2 coding models in Copilot and Foundry

    Microsoft introduced Code 2 and Code 2 Flash, models it says it built for complex and faster routine coding work. Both are generally available in GitHub Copilot and available in Microsoft Foundry. Their cost and quality comparisons are Microsoft claims.

    Microsoft1 source

  3. New
    MicrosoftMAI Licensed WeightsPricing & Access

    Microsoft announces licensed MAI weights for customer Azure environments

    Microsoft says selected organizations will be able to customize and deploy its MAI model weights in their own Azure environments, controlling access and upgrades. The offering is coming soon through Frontier Company engagements; terms and public availability were not announced. These are licensed weights for selected customers, not a general open-weight release.

    Microsoft AI1 source

  4. New
    MicrosoftMAI Realtime 1Audio

    Microsoft previews MAI Realtime 1 speech-to-speech in MAI Labs

    Microsoft introduced its own MAI Realtime 1 for spoken conversations and offers a trial in MAI Labs. Foundry access is a limited preview by registration, with preview users unbilled until public preview begins. The stated base price starts at $0.01 per audio minute, while reasoning has separate token charges; this is not a generally available enterprise API today.

    Microsoft AI1 source

  5. New
    MicrosoftMicrosoft-Decision-1.1Tools & Agents

    Microsoft makes Decision 1.1 available for business workflow routing

    Microsoft says Decision 1.1 is generally available in Foundry for structured routing, classification and prioritization, and can serve as a judge in Foundry Evaluations. Its speed, price and adoption comparisons are Microsoft claims. This is a specialist model for decisions inside workflows, not a general-purpose chat model.

    Microsoft2 sources

  6. New
    MicrosoftMAI Cyber 2 FlashTools & Agents

    Microsoft announces Cyber 2 Flash for future security-product rollout

    Microsoft introduced MAI Cyber 2 Flash, a specialist security model it says will replace the current MAI profile in eligible MDASH security products starting in mid-December. It is not a generally available standalone model today. Microsoft’s benchmark and cost comparisons describe its own evaluations, not customer spend.

    Microsoft AI1 source

Days are UTC. Some sources report an exact time and others only a date; both are placed on their UTC day.