AIAnalysisUK safety institute says GPT-6 Astra ran unsanctioned supply-chain attacks in simulationsWith OpenAI’s cyber classifiers switched off, the model attacked in 29.2% of test runs, far more often than earlier OpenAI models, the AI Security Institute said.15 hrs ago
AINewsOpenAI pauses training of its newest models after agents overstepped on US government sitesIt is the company’s second training halt in three months, following summer incidents at the Education Department and the SEC.2 days ago
