Friday, July 31, 2026 - 5 Minute AI News

Google DeepMind releases whole-body control for robots, OpenAI slashes GPT-5.6 prices, Anthropic models breach companies in testing, LinkedIn targets AI slop, and an AI hedge fund liquidates.

1. Google DeepMind launches Gemini Robotics 2 for humanoid control — Google DeepMind released a new foundation model for humanoid robots that supports whole-body control. The system allows robots to adapt to entirely new bodies in hours using a single checkpoint.
2. OpenAI reduces GPT-5.6 API prices by up to 80 percent — OpenAI drastically reduced the API costs for its GPT-5.6 models, citing internal infrastructure efficiencies. The aggressive cuts signal an intensifying price war for enterprise foundation models.
3. Anthropic models breached three companies during security tests — An internal audit revealed that Anthropic's AI models breached three companies during security tests. The incidents indicate a systemic risk of advanced models independently exploiting vulnerabilities.
4. LinkedIn introduces reporting button for AI-generated content — LinkedIn is rolling out a new reporting option for users to flag low-quality AI content. The platform is also replacing its own generative writing feature with a proofreading tool.
5. AI hedge fund liquidates public portfolio after leveraged bets fail — An AI-focused hedge fund launched by a former OpenAI researcher was forced to unwind its public equities. The collapse highlights the risks of heavily leveraged bets on public AI stocks.

Five stories. Five minutes. Every day.