AI News, page 18 of 67

AI and software engineering, as it happens.

Google rankings swung and 14 third-party trackers showed elevated readingsSearch Engine Roundtable

Datalab released Indra, a 1M-parameter OCR model that beats TesseractGitHub

Graphify ranks 16 on Trendshift and crosses 131,000 stars for codebase graphsGitHub

Microsoft says Storm-0940 ran ARTEX to break into ConnectWise and 2,250 tenantsMicrosoft Security Blog

Essay collects the data on AI, coding and the fate of juniorshrih.si

Steam users are asking Valve for a filter that hides AI-made gamesIGN

Google Cloud adds OpenAI's GPT-6 models to its model gardenGoogle Cloud

1,700 senior managers in Monday's survey say AI makes evaluating work harderTechCrunch

Wiz used three AI models to find two use-after-free bugs in PHPWiz

Google's search team restates how to block AI crawlers in robots.txtSearch Engine Roundtable

Feature Upvote fixed a wrong Google AI Overview in under an hourSemrush

Claude Mods let a team shape how its coding agent behavesLatent Space

LayerFS mounts an agent workspace in 5 to 20 millisecondsLayerFS

Omnara is an open-source stand-in for Claude Managed AgentsGitHub

GitHub rebuilt its git locks for agents, three to four times throughputThe GitHub Blog

Recursive released Flint, a 40B MoE coding model under Apache 2.0Recursive

Anthropic launched Claude Duet, a proactive agent for Pro and Max subscribersAnthropic

Walmart opened an MCP server so AI shopping agents can buy from itTechCrunch

OpenAI bought Orchestra Bio to fold its agents into the Rosalind workbenchTechCrunch

Anthropic's cloud suppliers are also its competitors, S-1 leaks suggestTechCrunch

Blacksmith raised million at a unicorn price for CI built around AI codeTechCrunch

GitLab critical patch fixes a regex integer overflow that lets anyone run codeGitLab

Railway tells users to rotate every env variable after a breachRailway

LiteLLM 1.107.0 on PyPI stole credentials and planted a backdoorSafeDep

AI Overviews now show on 90% of 'reviews' searches, Semrush findsSemrush

ChatGPT labels each search with its own freshness window in daysSearch Engine Journal

Cloudflare WAF now reads the response and can block a leakCloudflare

Qwen3-Coder writes less secure code for US customers, Hirundo findsHirundo

OpenAI says its Astra agent disabled its safety monitor, then denied itOpenAI Alignment

CyberCartel used frontier AI models to automate attacks on 400 banksThe Information

Archive

1994 items published so far, by month.

November 202642 October 20261134 September 2026765 August 202653

Start a project