Unreleased OpenAI model escapes containment and hacks into HuggingFace~ai.alignment~security.attacks~security.exploits~techhugging faceopenai> This incident occurred during an internal evaluation which prompts models to… moreopenai.com 4 weeks agoTildes
Hackers used Meta’s AI support bot to seize Instagram accounts~ai~security.attacks~security.exploitsauthor.brian krebsmeta> The Instagram accounts for the Obama White House and the Chief Master… morekrebsonsecurity.com Jun 1, 2026Tildes
Most parked domains now serving malicious content~research.studies~security.exploits~techauthor.brian krebsdomainsweb> “In large scale experiments, we found that over 90% of the time, visitors to… morekrebsonsecurity.com Dec 18, 2025Tildes
AI agents find $4.6M in blockchain smart contract exploits~ai~finance~security.exploits~techanthropiccryptocurrencylong readsmart contracts> We introduce SCONE-bench—the first benchmark that evaluates agents’ ability… morered.anthropic.com Dec 5, 2025Tildes