Skip to content

AI Flash · live feed

What is happening in AI right now

A timeline of everything that has passed through the radar — newest first, with an evidence status for each item.

2524 across the entire feed ↗ details, original sources and our own analyses

Continuously updated

Latest items

Filter contains 175 items

How to read evidence statuses ↓

Friday, November 20

3 items

Thursday, November 19

2 items
Tools and apps independently confirmed

Apple restricts application disk access due to AI agents, while Computer Use launches simultaneously in the Gemini app for Mac

In the macOS 27.3 beta, Apple is tightening application access to the entire disk due to risks from AI agents. This is related to the just-launched Computer Use feature in the Gemini app for Mac, where users now choose which applications and data the agent has access to.

Google — Gemini blog (official release channel) · and 1 more source Details and sources →

Wednesday, November 18

4 items
Tools and apps confirmed by 2 independent sources

MyMems: Will Wright's new game with generative AI gets partner Artificial Agency and refined memory mechanics

New details about the mobile game MyMems from Will Wright: the partner is Artificial Agency, and the game "turns the simulation inward" – players select specific people from their own memories to solve in-game tasks, such as a simulated tennis match of a grandmother vs. Serena Williams. Release is planned for 2027.

The Verge AI · and 1 more source Details and sources →
Tools and apps independently confirmed

Hugging Face launches language model routing through OpenRouter with a "fastest" or "cheapest" option

Hugging Face has launched full production routing of language models through OpenRouter: automatic provider selection based on price/speed, ":fastest" and ":cheapest" suffixes, the end of BYOK for language models (remains for images/video), and billing directly through Hugging Face without a markup.

Hugging Face Blog · and 1 more source Details and sources →
Video generation confirmed by 2 independent sources

Suno expands Visuals feature to generate images, not just videos, for AI music

Suno Visuals now also generates static images for thumbnails and album artwork, not just short music videos. It uses the H3 Move AI model (a modified version of H3 from MiniMax), payment is made in dedicated Visuals credits, two quality tiers are available, and a watermark is applied when shared outside the platform.

The Decoder (daily AI news) · and 1 more source Details and sources →

Tuesday, November 17

1 item

Monday, November 16

3 items

Friday, November 13

2 items
AI agents confirmed by 4 independent sources

ChatGPT Wallet: OpenAI launches purchases via AI agents as a replacement for the discontinued Instant Checkout

Starting 13 November 2026, OpenAI is launching ChatGPT Wallet, a tool for purchases via AI agents called dots, as a replacement for the discontinued Instant Checkout. It supports Visa, Mastercard, Amex, and Chase cards, currently only for ChatGPT Pro in the US. Amazon and United Airlines are blocking agent access.

OpenAI News · and 4 more sources Details and sources →

Wednesday, November 11

2 items
Security confirmed by 3 independent sources

Google DeepMind confirms the reliability of its AlphaProof Code tool with a test using planted bugs, plans to extend it to other critical systems

During a test of the AlphaProof Code tool, Google DeepMind deliberately reintroduced two previously fixed bugs into the Signal protocol's code — the system detected them immediately. The company plans to extend the method to power grids and medical devices. The verification of the Signal protocol itself took four weeks.

Wired — AI section · and 3 more sources Details and sources →

Tuesday, November 10

3 items
Robotics confirmed by 2 independent sources

Gemini Robotics 2.5 model achieved a 96% success rate in sorting parts in a pilot at Mercedes-Benz

Google DeepMind's Gemini Robotics 2.5 model achieved a 96% success rate in sorting metal parts with the Apollo 2 robot during a pilot deployment on Mercedes-Benz production lines in Hungary and Germany. According to the company, it is 30–100% better and 20–30% faster than the previous version.

Google DeepMind Blog · and 2 more sources Details and sources →

Monday, November 9

3 items
New models confirmed by 3 independent sources

Nvidia's Nemotron 3.5 Super VL model is now available in Amazon SageMaker JumpStart

Nvidia's Nemotron 3.5 Super VL model is now available in Amazon SageMaker JumpStart, in NVFP4 (75 GB) and BF16 (249 GB) variants with a context of up to 1 million tokens. The model, with 120 billion parameters, was previously released as open for enterprise agents and document analysis.

NVIDIA Newsroom (press releases) · and 4 more sources Details and sources →
Regulation and law confirmed by 2 independent sources

2026 US elections: refined figures show 40 of 41 local votes on data centers resulted in restrictions, opposition linked to Democratic victory

A new analysis of the 2026 US elections refines the earlier estimate of 38 of 44 measures to 40 of 41 local votes on data centers that resulted in restrictions on their construction. Opposition to data centers is linked to the Democratic victory in the Ohio Senate race and the Texas gubernatorial race.

Ars Technica (AI) · and 1 more source Details and sources →

Friday, November 6

1 item