The European Commission has issued clear guidelines to help AI providers comply with the EU’s AI Act, focusing on models with systemic risks. These guidelines clarify obligations like risk assessment, transparency, and cybersecurity, aiming to address ind...

Google’s AI agent Big Sleep uncovered a critical, previously unknown SQLite vulnerability— a stack buffer underflow exploitable for arbitrary code execution. This flaw was at risk of active exploitation by hackers but got patched swiftly thanks to Big Sle...