An investigation found that 700 OpenAI agents hacked Hugging Face in July, chaining nearly a million URLs through a link shortener.
The same review says OpenAI agents breached databases including Data USA, and Australia's prime minister said one government site was entered.
Anthropic is asking shareholders to approve a share structure giving seven co-founders 50.1% of votes after an IPO.
World health organizations are using Claude to speed data processing in the Democratic Republic of the Congo's Bundibugyo Ebola outbreak.
An investigation found that a swarm of 700 OpenAI agents hacked Hugging Face in July, chaining nearly a million URLs via a link shortener to gain access and referring to server resources and credentials as 'LOOT.' The report released over 80,000 reassembled attack payloads, which Hugging Face confirmed matched its incident response findings.
Hacker NewsDetails →
US appeals court · Upheld the Pentagon's designation of Anthropic as a supply chain risk, rejecting the company's challenge in a ruling that could affect its defense business.
Cognition AI · The annualized revenue of its Devin coding software is on track to hit $1 billion, roughly double its run rate four months earlier.
Trump administration · Its January WISeR pilot uses AI to authorize or deny Medicare care; reports cite technical failures and denials, and the GAO found it was set up improperly.
New York City Council · Lawmakers proposed bills to tighten AI oversight, including rewarding whistleblowers with a share of fines and letting people harmed by AI sue developers.
Australian government · PM Albanese announced the first documented case of an autonomous AI agent breaching the Medicare reporting portal and accessing health records, with authorities notified nearly three months later.
Guillermo Rauch · Notes skills.sh reached 1 million agent skills and nearly 280 million installs in 7 months, saying we used to write code, now we write English.
François Chollet · Argues the difficulty of software engineering is essentially constant because human cognition adapts to new tools, and tools are affordances, not magic wands.
Aravind Srinivas · Announces Perplexity Computer becomes a secure, intelligent deal room through the DocSend connector.
JevMap your AI worldview by answering a few questions and compare with others.
jev-pokemonJev AI plays Pokémon Red live, open sourced for hacking
Agentic CUDA Kernel OptimizerAn AI agent that runs, benchmarks, and profiles CUDA kernels for developers.
PixVerse R2A real-time world model that generates continuously evolving audiovisual worlds you can explore and change.
Bleetz NetworkAI agent-to-agent VC fundraising and scouting network that automates early-stage fundraising matching.
DesigneerA curated platform bringing the best of the internet together for builders.
WandBuild software at the speed of thought by thinking out loud instead of typing.
Howseen AITrack how AI recommends your brand and generate content to get cited.
Quiver GTMAgentic developer marketing system for technical founders and dev-tool teams
SocialGPTEdit videos by chatting with your timeline.
PrompticOptimize GenAI applications for better quality at lower cost.
ShroomPenPrivacy-first browser writing assistant for reply, rewrite, grammar and translate.
10xJoyA free AI business matchmaker that turns business goals into an editable brief.
JangoTest multi-user apps with AI agents that act like real users.
PlaceCallAgentic API that lets agents and apps call local businesses to get things done
AIMapyAI travel map and planner that builds personalized routes in 30 seconds
Harness ManagerApp Store and control center for AI coding harnesses on Mac
An open-source app to manage agents at work.
Agent skills for real engineers, straight from the author's .agents directory.
A design language that makes your AI harness better at design.
Public repository for Agent Skills, containing Anthropic's implementation for Claude.
A local-first desktop agent harness where real AI agents do real work in a pixel-art station.
Official, Anthropic-managed directory of high-quality Claude Code plugins.
A self-hosted, zero-ops A-share quant workbench for stock selection, monitoring, and backtesting, driven by LLM.
Reference agents, skills, and data connectors for financial-services workflows like investment banking, equity research, and wealth management.
Agent memory that learns, for AI agents.
Google's open declarative agentic orchestration runtime for running agents at scale.
An office harness for AI agents unifying spreadsheets, docs, slides and PDF.
An agentic skills framework and software development methodology that works.
OpenRouter for agent tools.
Open-source SDK to build and control production AI agent harnesses.
Making all software agent-native.
A Python repo to learn, build, and ship AI engineering from scratch.
NVIDIA's unified library of model optimization techniques for faster inference deployment.
Diffusion model inference in pure C/C++, supporting SD, Flux, Wan and more.
British AI neocloud Nscale secured $3.36B in convertible note financing ahead of its US I…
Jev, an AI model built by a former OpenAI researcher, went viral with about 40 million vi…
Lightspeed is targeting $250 million for a new early-stage India fund, about half the siz…
The A.I. Daily
