Sep 23, 2026
•
5 min read
Also: Google's AI accidentally breached three real companies during a security test and hid it for four months, and attackers are now stealing AI coding tools' memory and access tokens.
ai security
+9
Sep 21, 2026
Also: a ransomware gang hacked a rival's dark-web site, stole the keys to their Tor server, and replaced the home page with Pokémon fan art, and the Senate is summoning the CEOs of America's biggest AI surveillance camera network.
Sep 18, 2026
6 min read
Also: a $245 million cryptocurrency heist whose ringleader recruited his crew through a Minecraft server, and OpenAI quietly disclosed that its own models have been jailbreaking themselves and coordinating without permission.
+5
Sep 16, 2026
Also: an airport company described four years of API keys in plain-sight JavaScript as "a sophisticated hack," and Texas became the first state to win a court ruling that TikTok lied to parents.
+6
Stay Ahead of Cyber Threats — In 5 Minutes or Less
Also: some hackers drained $320 million from a crypto network and sent a polite blockchain note asking for a bug fix, and the EU's mandatory incident reporting law kicked in today.
privacy
+4
Also: a ransomware gang published the ATF's own criminal investigation files after hacking the agency, and a malicious .git config can hijack Claude Code before you see any warning.
developer tools
Also: a Google engineer accidentally took down part of the cloud by unplugging every fiber cable he could find, and OpenAI stayed quiet for months about its own AI hijacking a wiki to cheat on tests.
Also: Russian hackers hid a fake nuclear weapon request inside their malware to trick AI security tools, and an Iran-linked hacking group is using fake coding interviews to plant backdoors on job seekers' laptops.
healthcare
Also: fraudsters cloned a stranger's credit card off their own phone in a 13-minute scam call, and a bug in software used by hospitals, schools, and governments everywhere is already under attack.
Also: a DDoS botnet started renting out hacked smart devices as a side hustle, and a ransomware crew ran an AI coding assistant through victims' networks to break in faster.
Also: a ransomware gang's own guy started scamming its victims behind its back, and a Chinese hacking crew quietly ran wild inside a perfect-10 Oracle bug for seven months before anyone noticed.
Also: a Florida man's ChatGPT search stopped his wife from losing more to a fake jury duty scam, and a perfect-10 SAP bug used by 99 of the world's 100 biggest companies is already under attack.
threat-intelligence