You're tuned in to Tech Beat. Here are the stories shaping the conversation today.
Microsoft's Mustafa Suleiman is claiming his company's AI system outperforms doctors on difficult diagnostic cases, calling it a step toward what he's calling medical superintelligence. A study published in Science backs the performance gains, though researchers note the credit likely belongs to the underlying AI models rather than Microsoft's own architecture. The distinction matters, because it shapes how we understand who's actually driving progress here.
On a different frontier, Anthropic put its Claude model to work as a security auditor, running it across twenty-five major open source projects. The results were striking — nearly five hundred vulnerabilities flagged across C, Java, and Python codebases, including one flaw that had gone undetected for twenty-one years. It's a reminder that AI's most unglamorous applications may also be its most consequential.
And in orbit, Britain is putting serious money behind the space debris problem. The UK Space Agency has awarded a contract worth roughly eighty-one million dollars to retrieve two defunct satellites between twenty twenty-nine and twenty thirty-two, using manipulator spacecraft launching from the Shetland Islands. It's a small mission by space standards, but someone has to start cleaning up.
That's your briefing for today. Keep surfing. Tech Beat out.
