Fresh bugs in a Microsoft BitLocker security wrapper highlight how compromise risk can begin in the software around encryption, not only in the encryption engine itself.
Named flaws in two popular Chrome extensions underline how an AI helper with broad page access can turn convenience into high-impact compromise risk.
A newly flagged ChromaDB weakness underscores a hard truth in AI infrastructure: if request handling and trust checks are ordered badly, an ordinary API call can become a code-execution event.