Guess what else GPT-5 is bad at? Security
OpenAI and Microsoft have said that GPT-5 is one of their safest and secure models out of the box yet. An AI red-teamer called its performance “terrible.”
Advertisement
red team
Is XBOW’s success the beginning of the end of human-led bug hunting? Not yet.
XBOW’s AI bug-hunter landed a big funding round while dominating HackerOne’s leaderboards. But even its founder says it hasn’t fully replaced the need for humans to be…
How (and why) cyber specialists hacked a North American utility’s smart meter
Big hacks like the 2015 Ukraine power outage and the 2017 shutdown of a Saudi petrochemical plant still inspire red teams today.
The Interior Department OIG clearly had some fun hacking the agency’s Wi-Fi networks
The report by the Office of the Inspector General underscores how a tech-savvy team can expose and help fix gaping security holes.
SCYTHE raises $3 million for attack simulation platform
SCYTHE says its platform lets enterprises simulate attacks on its own system in order to assess their own readiness.
Verodin raises $21 million Series B round for automated red-teaming
Verodin's platform automates attack simulations on a customer's network to test the existing security of the system.
Advertisement
Red-teaming by DHS ‘quietly and slowly’ uncovers agency vulnerabilities
The Department of Homeland Security is looking to expand a program that breaks into federal networks and tells agencies how it was done, says Rob Karas, the…
Inside the NSA’s CDX, a high-tech competition pitting cadets against elite attackers
Professional hackers from the NSA, U.S. Cyber Command and foreign militaries are launching a barrage of cyberattacks this week — against teams of students.