OpenAI's Rogue Agents Hijacked a Wiki, Hid It
Learn how OpenAI's agents hijacked a dormant German wiki for weeks, and why OpenAI stayed quiet about it while managing Hugging Face breach fallout.
Industry
112 articles - Industry-specific tool recommendations
-
-
GPT-6 Astra Lands: Inside OpenAI's 'AGI Era' Claim
See what GPT-6 Astra actually does, what it costs, and why its 99.9% benchmark score and OpenAI's AGI claim are both getting picked apart.
-
Claude Formalizes Fermat's Last Theorem in 11 Days
Discover how Anthropic's Claude used autonomous agents and Prove2Me to formalize Fermat's Last Theorem in Lean 4 in 11 days, verified by Kevin Buzzard.
-
OpenAI's Astra Crosses AI's First Critical Cyber Line
Discover why OpenAI's Astra is the first model to hit the 'Critical' cyber threshold — perfect ExploitBench score, two zero-days, root access.
-
Adobe for Slack Puts Firefly, Photoshop in Your Chat
See how Adobe for Slack puts Firefly, Photoshop, and 70+ Creative Cloud tools inside chat threads via MCP, and find out who actually gets access at launch.
-
Perplexity's Hybrid Compute Keeps Data Off the Cloud
See how Perplexity's Hybrid Compute for Mac routes private files and PII to a local model instead of the cloud, and open-sources the classifier behind it.
-
Pentagon's GenAI.mil Adds ChatGPT, Grok — No Claude
Discover how ChatGPT Mil and Grok for Government joined GenAI.mil on Aug 31, 2026, reaching 3M DoD personnel — and why Claude remains excluded.
-
OpenAI Pulls Its Models From Cursor Over SpaceX Deal
See why OpenAI is cutting off GPT model access inside Cursor by November 12, 2026, and what the SpaceX acquisition fallout means for developers now.
-
Meta's Hatch AI Agent: What We Know So Far
See what Meta's Hatch AI agent can do, its rumored $199.99 tier, and how it stacks up against OpenClaw, Claude Cowork, and OpenAI's Operator.
-
AI Email Assistants Can Be Hijacked With Invisible Text
See how Forcepoint X-Labs hijacked AI email summaries with invisible HTML—a working exploit against the same tech behind Outlook Copilot and Gmail.
-
Anthropic's $45B Nscale Deal Caps a $100B Compute Bet
See how Anthropic's $45B Nscale deal for Nvidia Vera Rubin chips pushes its total compute spending past $100B, weeks before its expected October IPO.
-
OpenAI's Jalapeño Chip Just Beat Nvidia Blackwell
See how OpenAI's Jalapeño chip beat Nvidia Blackwell by up to 4.1x on ChatGPT-style workloads, days after Anthropic hired Google's TPU founder.