Category / AI

AI

Changes in AI models, capabilities, and agents, clearly explained.

News & changes

New announcements, clearly explained

Newest first publication

Check who a change applies to, when it starts, and the conditions confirmed by official sources.

What changed

Edge Aion AI: site trials and device requirements

Microsoft is replacing Edge's built-in model with Aion and opening a live-site Prompt API trial. We connect model availability with site registration, device requirements and the scope of local processing.

AI 10 min read
What changed

Gemini Mac computer use: U.S. rollout and controls

Gemini can handle clicks and typing across Mac apps and the web. We connect the separate working window and user controls with region, language, hardware and account requirements.

AI 6 min read
What changed

Copilot Adds Eight Connectors: Access and Admin Rules

Copilot adds eight external service connections in public preview across all plans. Company accounts must meet organization and service access requirements.

AI 10 min read
Now

Gemini 4 Neon: Pro and Ultra Access, Thinking Controls

Gemini 4 Neon is rolling out to Pro and Ultra subscribers. App model choices and usage rules need to be read separately from API pricing and evaluation settings.

AI 7 min read
What changed

Genie 4 Rolls Out: Ultra Access, Pro Waitlist and Sharing

Genie 4 is rolling out to eligible Ultra subscribers. Pro users can join a waitlist; allowing remix when sharing can expose the original prompts and images.

AI 7 min read
Now

OpenAI Chip Design: What It Verified on Silicon

OpenAI’s November report details AI-proposed changes to its Jalapeño chip and to AMD designs. A separate RISC-V exercise estimated throughput in simulation.

AI 4 min read
Now

Gemini for Android Go: 2 GB RAM Minimum

Gemini replaces the previous assistant on eligible Android Go devices. It requires at least 2 GB of RAM, and the way to open it depends on the device.

AI 4 min read
Now

Google Stitch Loop: Waitlist and Human Review

Google Labs opened a global waitlist for Stitch Loop, an experiment for maintaining live products. It can prepare changes, but a team still decides what merges.

AI 4 min read
Now

Meta Muse Adds Loblaw Grocery Checkout in Canada

Meta’s AI assistant Muse now connects with Loblaw in Canada. At participating stores, shoppers can move from choosing a store to building a basket and checking out inside Muse.

AI 4 min read
What changed

Meta Muse Launches in Australia, With Local Connectors

Meta’s Muse is available in Australia on mobile, web and Mac. Local connections are arriving in stages, while Canva is a connector across supported Muse markets.

AI 4 min read
What changed

Microsoft Copilot Autopilot Starts US Frontier Preview

Microsoft is rolling out Copilot Autopilot in preview. Here’s who qualifies in the US and how the agent handles approval before acting.

AI 5 min read
What changed

Google Drive Adds Testing for Gemini Auto-Labeling

Gemini auto-labeling in Drive is now generally available, with a way to test instructions before turning them on and reporting to monitor the results.

AI 5 min read
Now

GitHub Copilot App Previews Azure DevOps Integration

Teams can now work with Azure DevOps code and pull requests in the GitHub Copilot app. Here’s what the preview means for permissions and data.

AI 4 min read
What changed

Copilot Fusion Adds Web, Visual Studio and Mobile Beta

GitHub Copilot Fusion reaches more places. Here’s what changed on the web, in Visual Studio and on mobile, and how using more models affects billing.

AI 4 min read
Now

Gemini Tasks Starts Rolling Out Free in South Korea

South Korea is one of six markets getting Tasks in Gemini without a subscription. Here’s what accounts qualify and which features start in the US.

AI 4 min read
Now

GPT-6.2 Astra and Sol: Access and API Prices

GPT-6.2 Astra and Sol are available in ChatGPT Work and Codex. Here’s how regular access, the faster option and standard API prices differ.

AI 5 min read
Now

Tripo Previews H3.5 and Polipop for 3D Printing

Tripo showed objects generated with H3.5 and demonstrated Polipop, a tool for taking an idea through print preparation and on to a printer.

AI 4 min read
Now

OpenAI Details NATO’s Use of Defensive AI

OpenAI says it gives core NATO institutions access to defensive AI models. The monthly token figure it released measures how much the models processed.

AI 4 min read
What changed

Gemini Skills Can Be Shared, With Separate Drive Access

Gemini’s web app can share reusable instructions called skills. Viewers can use a skill, while linked Drive files still need their own access permissions.

AI 4 min read
What changed

Google Pics Upgrades to Nano Banana 2.1

Google Pics moves to an image model designed to render lettering more accurately and preserve more of the original during edits. The upgrade also reaches Pics tools in Docs and Slides.

AI 4 min read
Now

ChatGPT for Healthcare Adds Agents With Review

OpenAI announced agents that can work across healthcare screens, improved evidence search, and new healthcare plugins. Organisations and verified U.S. clinicians have different access routes.

AI 5 min read
Now

OpenAI’s Science Blog Opens With an AI Physics Case

OpenAI opened a blog about scientific results and how researchers use AI. The first guest post describes GPT-6 Astra’s role in developing a mathematical proposal for the Standard Model.

AI 5 min read
What changed

GitHub Copilot to Retire Six Models on Dec. 14

GitHub plans to retire six Copilot models on December 14. Haiku 4.5 stays for individual annual subscribers; the Kimi K3 alternative is still pending.

AI 5 min read
Now

Gemini 4 Hydrogen: Free App and Search Access

The free Gemini app can choose Hydrogen automatically when it suits the question. Google Search and Sheets offer the model under different conditions.

AI 5 min read
Now

Gemini 4 Carbon: limited access and test conditions

The October 28 announcement set out selected-defender access and security test conditions. Its access plan is anchored to that announcement date.

AI 8 min read
What changed

Workspace skills: sharing copies and access

Sharing a skill creates an independent copy; access to referenced files is separate. Workspace skills also stay separate from skills and Gems in the Gemini app.

AI 10 min read
What changed

Copilot Fable 5.2: admin access and data retention

Availability, organization policy and retention approval are separate checks. Eligible enterprises still need an administrator to enable the model after ZDR approval.

AI 6 min read
What changed

Claude Fable 5.2: access and one-week usage limits

Fable and Mythos share a model with different access and safeguards. Compare the one-week Pro and standard Team offer with what a Fable reset restores.

AI 9 min read
What changed

K-EXAONE 2.5 API opens free for noncommercial use

Free API access enables experimentation, with noncommercial-use terms, restrictions on third-party model access, and choices about training use of inputs.

AI 6 min read
What changed

Claude Enterprise Tags: Which AI Spend Gets Counted?

Claude Enterprise can group AI spend using codes on projects. Spend before a tag takes effect and usage included in a seat do not count toward the new tag.

AI 6 min read
What changed

GitHub Copilot Shows Which Files It Didn’t Review

GitHub Copilot now groups skipped and excluded files in its review overview. The management API can also identify Copilot output and expose more settings.

AI 6 min read
What changed

Premiere Beta AI Assistant Switches to Firefly Credits

The Premiere beta AI Assistant now draws on your Adobe plan’s Firefly credits. You can see what a request used once the assistant responds.

AI 6 min read
What changed

Meta Muse Spaces: Display Schedules and App Updates

Build a reusable screen inside Muse and pin it to a tab. When the screen appears is separate from when connected app data refreshes.

AI 5 min read
Now

Kakao’s Kanana AI Returns Choice Probabilities

Kakao released an AI model that reads text and photos and returns probabilities or scores. A support example shows what it can decide and what the license covers.

AI 6 min read
Now

Adobe Coworker in ChatGPT and Claude: Who Can Read What

Adobe Coworker brings company marketing insights to ChatGPT and Claude. Administrators approve access, and users can read only the Adobe data their own permissions allow.

AI 5 min read
Now

GPT-6 Cyber: Who Gets Access and What Tests Show

GPT-6 Cyber starts with separately approved Daybreak Red organizations. OpenAI’s tests of cyber capability and staying within assigned scope measure different things.

AI 5 min read
Now

Claude Personal Finance: What Remains After Unlinking

Claude’s US finance beta reads linked accounts to explain spending. Unlinking a bank stops new reads, while earlier answers and shared outputs can remain.

AI 6 min read
What changed

Meta Muse Voice: US Rollout and Data Controls

Meta is rolling out live voice for Muse in the US. Users can choose audio or an animated character, while a separate setting governs use of interactions for model improvement.

AI 5 min read
Now

Gemini Omni 1.2: 10s vs. 15s API Clips and Cost

Omni 1.2 Flash can make clips up to 15 seconds. The choice between 10 and 15 seconds changes how many reference clips you can use and what 720p output costs.

AI 5 min read
Now

Copilot in Xcode: Built-in Preview vs. Extension

GitHub Copilot now has a preview inside Xcode’s coding assistant. Compare the required versions, setup and permissions with the separate extension.

AI 5 min read
Now

Siri AI App Integration: Who Can Test the Preview?

Apple has opened testing for custom app actions in Siri AI. Here’s what developers can expose and which copies of an app can try the preview.

AI 5 min read
Now

SKT AX 2.0 shifts from AI tools to workflow redesign

SKT is expanding internal AI use toward organizational workflows. Its three support programs and data tools explain the plan, while adoption counts describe its current reach.

AI 7 min read
Now

Genie 4: interactive worlds, audio and preview access

Google DeepMind has started a trusted-tester preview of Genie 4. We explain interactive scenes and audio, consistency limits, and future Ultra access.

AI 10 min read
Now

GPT-6.1 Terra: screen tasks, access and API costs

Connect GPT-6.1 Terra access with API billing, and distinguish an illustrative request price from the cost of a completed workflow.

AI 8 min read
Now

Decisions API: answer types, costs and data controls

The generally available API returns probabilities, choices or scores for questions an app defines. Its results, actual actions, input costs and data controls serve different roles.

AI 7 min read
Now

OpenAI model gateways: approval rule starts November 20

The policy distinguishes direct model access from model integration within app features. We connect the review requirement, delivery roles and effective date.

AI 5 min read
Now

OpenAI’s AI shutdown proposal: who controls what?

Who can interrupt an AI agent’s work? OpenAI’s policy proposal connects each actor’s control boundary with monitoring, intervention and safe resumption.

AI 4 min read
Now

OpenAI plans Google Cloud model launch in coming weeks

General availability is planned in the coming weeks. Model discovery, Marketplace transactions and use of existing Google Cloud commitments are separate parts of the plan.

AI 5 min read
What changed

Kakao releases two Kanana vision encoders

The released models extract image features for language models. Their evaluation results and commercial terms need to be read in the context of the task and deployment.

AI 9 min read
What changed

Clairic preview separates extracted and estimated values

Clairic returns numbers, dates and other typed values for workflows. Extracted facts and predictions about unknown outcomes call for different evidence.

AI 8 min read
What changed

GitHub adds HydraFusion Max and puts more reasoning into review

HydraFusion now applies different reasoning effort to drafting and review, while Max targets more demanding work. Its workflow changes come with preview-setting and organization-policy conditions.

AI 8 min read
What changed

GitHub Copilot Adds Mandatory Policy Refresh at Startup

Enterprise administrators can require Copilot to confirm current managed settings before startup and restrict clients when that confirmation fails.

AI 8 min read
What changed

GitHub Copilot in Slack and Teams: Who Can Use It?

GitHub says @GitHub is generally available in Slack and Teams. Teams adds a canvas where people can review plans and code changes beside the conversation.

AI 7 min read
Explanations & how to

Understand, compare, and use

Newest first publication

Explanations and practical steps for questions you can return to. Check each story’s sources and dates for changing prices, availability, and device support.

Explained

How Codex investigates ChatGPT slowdowns

OpenAI's Performance Factory uses Codex to investigate performance problems and propose fixes. We explain what its investigation counts and review sample measure, and how human review connects with recovery checks.

AI 9 min read
Explained

OpenAI Control Room: what the 35-second stop covers

The timer starts with a stop instruction. Research jobs can remain alive while output is blocked; detection, decisions, and intervention need separate scrutiny.

AI 4 min read
Explained

GR00T 2’s 95% success rate: the conditions behind it

Success at least once on 40 tasks differs from average success across attempts. GR00T 2’s demonstrations, fine-tuning and code timeline explain the result.

AI 8 min read
Explained

OpenAI Lab Benchmark: What the 66.7% Score Measures

GPT‑6 Astra averaged 66.7% of available credit across 100 tasks. The number measures how it analyzed evidence, rather than how often experiments worked.

AI 5 min read
Explained

OpenAI training security: isolation, alerts and stopping

Isolation, monitoring and stopping play different roles in AI research security. We connect offline browsing and an earlier incident with the case for restarting.

AI 6 min read
Explained

MAI-DxO study: accuracy, test costs and limits

Microsoft's MAI-DxO results bring accuracy, information gathering and estimated test costs into one evaluation. We explain its rules, data access and planned clinical assessment.

AI 7 min read
Explained

TK Elevator AI: connecting sensor data and service notes

Sensors show recurring signals, while technicians’ notes add what was found on site. Microsoft’s case connects pre-visit preparation, human judgment and spoken reports that inform the next maintenance visit.

AI 9 min read
Explained

ChatGPT Search vs. Google: What Actually Changes?

ChatGPT Search turns information from the web into a conversational answer. It is useful for a quick overview, but important decisions still require checking the linked source, its date, and its scope.

AI 7 min read