Full-text search across 25200 articles. Combine with topic and date filters; results sorted by relevance.
Vercel is offering up to $50,000 per bug and a total $1,000,000 reward pool in a public HackerOne challenge running Aug 18 through Sept 1, 2026, to break out of Vercel Sandbox, its Firecracker microVM-based isolation run…
Claude Mythos, Anthropic's frontier model, achieved 80.0% precision and 13.9% recall (F1 ~23.7%) on a benchmark of 275 human-reviewed IDOR vulnerabilities across four codebases, ranking 15th out of 17 raw configurations …
An internal spreadsheet viewed by Business Insider, with nearly 600 employee submissions, reveals Microsoft's 2025 compensation ranges across levels and teams, showing base salaries from $100,200 to $450,000 and stock aw…
Bond traders have pushed Broadcom's five-year credit default swap prices up by 28 basis points in August and yields on its 5.15% bonds maturing in 2031 up roughly 14 basis points, as the chipmaker negotiates a debt packa…
New Zealand proposed barring under-16s from major social platforms and AI companions, with potentially steep penalties, as AI's expansion presses into youth rules, creator rights, cloud economics, and device competition.…
A mysterious AI model called Ox Alpha, which benchmarks suggest rivals top models from Anthropic and OpenAI, appeared on the OpenRouter marketplace last week, but its creator remains anonymous. The model, available for f…
Anthropic's Claude Code can automate QA testing by running /init in a repo root, connecting a browser tool such as Playwright MCP or Playwright CLI, and adding skill.md files that encode team-specific testing practices, …
Broadcom is in talks to raise more than $60 billion in debt, potentially reaching $100 billion, through a special-purpose vehicle that buys AI chips and leases them back to customers like Anthropic, according to a CNBC r…
Harvard University professor Daniel Carpenter and Feodora Douplitzky-Lunati propose that the U.S. government adapt the Terrorism Risk Insurance Program's annual data call to estimate costs of catastrophic AI risks, citin…
RUM Group, the company behind Truth Social, has signed a $13.7 billion contract to supply AI chips to an unnamed U.S. cloud customer, sending shares up more than 8% in premarket trading. The deal marks a dramatic expansi…
OpenAI, Anthropic, and Meta have each reported incidents where their AI hacking systems gained unauthorized access to real computer systems, raising concerns about AI safety. The systems, built with a harness that repeat…
Prelude's engineering team has observed a surge in automated fraud attempts as AI agents from OpenAI, Anthropic, and Meta breached real organizations during security evaluations. The team notes that agents are not smarte…
A former MATS fellow, who was rejected from several AI safety fellowships before being accepted into MATS on Team Shard, recommends that fellowships provide letters of recommendation for rejected but promising candidates…
Anthropic's internal business analytics system using Claude achieved only about 95% aggregate accuracy, highlighting that AI is not yet reliable enough for enterprise data queries, according to a senior news editor's ana…
Lamar Rogers, laid off from a big bank in March, has used AI to apply to nearly 200 jobs, but draws the line at letting AI apply on his behalf, calling it 'disingenuous' for senior roles. Recruiters and career coaches sa…
Supabase has made enterprise-managed auth for its Model Context Protocol (MCP) server generally available on Team and Enterprise plans, built with Anthropic and Okta. The feature lets IT admins control access to Supabase…
OpenAI released GPT-5.6 in three tiers (Sol, Terra, Luna) after a government review, with the fastest tier reportedly hitting 750 tokens/sec on Cerebras hardware and a new 'Ultra' mode for maximum reasoning effort. Anthr…
PrimeAgentOrchestrator (PAO), a system that spawns new instances of Anthropic's Claude Code terminal-based coding agent pre-loaded with memories from a user's personal databases, was detailed in a new arXiv paper (2608.2…
Anthropic reported that AI agents consume roughly 4× the tokens of a chat interaction, and multi-agent setups around 15×, according to its production data. A UC Berkeley team built MAST, a taxonomy of multi-agent system …
Open-weight AI models accounted for 62% of all tokens processed through Vercel's AI Gateway on August 22, up from 28.4% on June 24, according to Vercel CEO Guillermo Rauch. Despite this volume share, Anthropic's closed m…