Latest cybersecurity reporting from selected sources.

Anthropic's Mythos Leak Points to Pattern of Failures, Sloppy Practices at AI LabsAnthropic accidentally exposed its most powerful unreleased AI model to compromise, and days later shipped its flagship coding tool's full source code without meaning to. Meta, Microsoft and OpenAI have each had comparable moments. Questions linger about the integrity of third-party AI tools.

Company Officials Reportedly Said Hackers Just Obtained Read-Only Access to CodeRussian nation-state hackers who compromised Microsoft's source code repository gained read-only access but not the ability to change code, top company officials reportedly told a German parliamentary committee on Wednesday. Microsoft is being criticized for high-profile security failures.

Warnings: Products' Source Code at Risk, Customers' Secrets Spilled via EmailWhat did Microsoft mean when it said that nation-state hackers are "attempting to use secrets of different types" they found in its emails with customers, what's the risk posed by those attackers also accessing Microsoft's source code and what can - and should - customers do about it all?

Russian Foreign Intelligence Service Hack Gets Worse for Computing GiantA Russian state hack against Microsoft was more serious than initially supposed, Microsoft acknowledged in a Friday disclosure to federal regulators. Microsoft said a Moscow threat actor obtained access to "source code repositories and internal systems."

Computing Giant Says Hackers Did Not Access Customer Data or Production SystemsRussian state hackers obtained access to the inboxes of senior Microsoft executives for at least six weeks, the computing giant disclosed late Friday afternoon. "There is no evidence that the threat actor had any access to customer environments, production systems, source code, or AI systems."

Computing Giant Says Hackers Did Not Access Customer Data or Production SystemsRussian state hackers obtained access to the inboxes of senior Microsoft executives for at least six weeks, the computing giant disclosed late Friday afternoon. "There is no evidence that the threat actor had any access to customer environments, production systems, source code, or AI systems."