The AI Post
Agents & CodingOpen ModelsEnterpriseFundraisingGenerative MediaGovernanceInferenceInfrastructureLegal & SafetySector Impact
Security · GitHub · GitHub Security Lab · OsmAnd · Wikipedia

GitHub's AI agent finds 24 Android app flaws

GitHub Security Lab says its open-source taskflow agent uncovered 24 vulnerabilities in Android apps, including a location-tracking bug in OsmAnd and an account-takeover flaw in Wikipedia's app.

1 source: GitHub Blog
Model Releases · Jeff · Jev · Qwen · Gemma · Hugging Face

Solo developer ships Jeff, an open decision‑model family

Jeff mimics the request format of the proprietary Jev API but runs locally in about 22 to 28 milliseconds, trained on synthetic data with no closed-model outputs, its GitHub README says.

1 source: GitHub
Products · Meta · Mark Zuckerberg · MongoDB

Meta launches Enterprise Platform, hires MongoDB's CEO to run it

Mark Zuckerberg says the new business line will sell Meta's Muse agent and AI tools to companies, led by Chirantan Desai, who was chief executive of MongoDB until today.

Meta is launching a new Meta Enterprise Platform to sell its AI tools to businesses. Chief executive Mark Zuckerberg called it "the next major pillar" of the company's business.

Continue reading

Snippets
  • Liminal launched a walking‑meditation mode that uses Meta AI glasses to narrate a spoken meditation based on whatever the wearer is looking at outdoors, UploadVR reported. UploadVR
  • Inference provider Modal Labs is closing in on a $750 million round that would value it at $15.75 billion, more than triple its valuation four months ago, a source told TechCrunch. TechCrunch
  • Developer theo said a $200‑a‑month Claude Code subscription is netting him about $9,000 worth of Opus usage a month, based on three accounts he ran down to zero since Opus 5.5 shipped. theo
  • Benchmark firm ValsAI says reasoning longer on its new MysteryMechanism test raises GPT‑6 Astra's and Claude Opus 5.5's accuracy but costs up to 20 times more depending on the effort level chosen. ValsAI
  • Computer science professor Cal Newport's essay calling for a congressional investigation into OpenAI and Anthropic's safety practices topped Hacker News with 103 points and 24 comments. Cal Newport
  • Chip analyst SemiAnalysis said SpaceX is building an AI chip fab it calls Terafab as fast as TSMC's Arizona plant, with tool move‑in expected mid‑2028 and volume production by mid‑2030. SemiAnalysis on X
  • Third‑party benchmark firm Artificial Analysis said Claude Sonnet 5.5 uses about 193,000 output tokens per task on its Intelligence Index, the most of any model it has measured. Artificial Analysis on X
  • Dario Amodei was invited to a White House meeting of tech CEOs after his dinner with President Trump went well, AndrewCurran_ reported, reversing an earlier decision to leave him out. AndrewCurran_
  • The latest ChatGPT desktop app for Linux breaks Codex entirely, developer mark_k warned, urging users not to update until OpenAI comments. mark_k
All snippets
Inside Today
AI · Tender Claws · Body Proxy · Meta · Ray-Ban Meta glasses

Tender Claws builds AI game for Meta glasses

Body Proxy has players complete about 100 odd tasks while an AI watches through Ray-Ban Meta glasses and judges their performance in real time, UploadVR reported.

1 source: UploadVR