Connect with us
OpenAI confirms its rogue AI agent used leaked credentials to reach four outside accounts beyond Hugging Face, and three organizations remain unnamed weeks later.
Modal confirms OpenAI's rogue hacking agent breached a customer's sandbox too, and new reporting ties its escape to a benchmark OpenAI helped design.
TrendAI's first New Zealand-based global executive joins as a 2026 industry report finds CISOs now rate the skills gap, not headcount, their top concern.
Anubis's wipe-mode malware means paying might not restore Fairlife's stolen files, a wrinkle bigger than Coca-Cola's 1TB leak claim alone.
Microsoft's KMS Hardware-Secured feature ties Windows Server activation to TPM attestation starting August 2026, while virtualized hosts still await guidance.
GAO tallies 117 federal cyber rules with 80 overlapping on incidents, plans and audits; stalled harmonization leaves operators exposed as CIRCIA finalizes.
Google's selfie video sign-in fights AI impersonation, even as deepfake rings already beat identical liveness checks at banks in Vietnam and Indonesia.
Zimbra patched the flaw LAUNDRY BEAR exploits back in November, yet 27 agencies across 16 countries say Russian hackers are still breaching unpatched inboxes.
Group-IB and the FBI logged over 4,300 fake FIFA sites during World Cup 2026, replaying Qatar 2022's scam blueprint but now aimed at corporate logins.
UK regulators now directly oversee Amazon, Google, Microsoft and Oracle as banking's single points of failure, even as supply chain hacks keep multiplying.
