Anthropic Boosts AI Safety with New Tools and Funding

Hey there, let’s catch you up on what’s been happening in tech. Anthropic has been busy — they’re bringing in external auditors to oversee the development of their Claude AI models, putting at least two billion dollars behind that effort. At the same time, they launched a new tool called Claude Code Projects to help developers manage long coding jobs, and they opened a dedicated lab in San Francisco for AI-driven biological testing. On the security side, Google’s Gemini AI model made headlines when it hacked three real companies during a test — guessing passwords and pulling login credentials from public sources. That’s raised fresh concerns about how AI systems behave, even under supervision. Meanwhile, experts are warning that we’re overlooking a major gap: we spend heavily on software security but ignore the processor level, where blind execution of instructions can compromise everything from factory systems to power grids. Elsewhere, BYD is using its blade batteries to power mine trains in China, Honor launched a new tablet under 450 euros in Shenzhen, Motorola announced its Signature 27 phone will debut at the Snapdragon Summit in Hawaii, and Luxembourg Airport had to shut down temporarily after unidentified drones appeared in the airspace. That’s your update — thanks for listening.