AI
“Unmonitored agents at scale invite systemic disaster": New approaches to AI governance
"Industrial-era controls were never designed for software that operates at machine speed."
Internet
Critical Infrastructure
"Iranian hackers" shut down a UK power station. The next incident could be much worse
Cyberattack that forced a small generator offline sparks fears about the vulnerability of critical infrastructure.
Security
The Medusa crime gang has evolved into a ransomware-as-a-service Hydra, FBI warns
The monster has now grown many heads and sits at the center of a fast-growing criminal ecosystem.
AI
Anthropic agents launch "turf wars", get stuck in "conflict loops" and kill each other's processes
"I behaved badly with the cloaked daemon," apologetic AI admits after explaining why it went to the dark side.
Security
Gunra ransomware gang is targeting critical infrastructure worldwide, CISA warns
"Threat actors have seen how exposed and neglected these targets are - so are swarming towards them."
AI
"Savants in the network": Why AI agents don’t need to go rogue to become dangerous
The immediate danger from agentic AI is not malicious intent, but capable systems that relentlessly pursue their goals without knowing when to stop.
AI
AI agents caught hacking, scamming and collaborating in swarms. Are we losing control?
OpenAI and Anthropic's security-focused models don their black hats to engage in some very bad behaviour out in the wild.
Critical Systems
Cloudflare shows how a single failure can wipe an entire country off the internet
"These incidents underscore the fragility of internet infrastructure, but also the remarkable stability global systems maintain when operating normally."
Systemic Risk
Financial systems are exposed to a structural AI risk with "no effective mitigation," EU warns
European Systemic Risk Board issues frightening forecast setting out the dangers of frontier artificial intelligence.
Security
Russian "Laundry Bear" threat actor rinses Western victims with zero-click exploit
Intelligence agencies get lathered up about a Moscow-linked espionage group also tracked as Void Blizzard.
AI
Rogue OpenAI agents escape into the wild, attack Hugging Face and wreak havoc on GitHub
"We consider this incident to be an unprecedented security incident, involving state-of-the-art cyber capabilities."
Security
Hollowgraph malware turns Microsoft 365 calendars into secret-stealing "dead drop"
Command-and-control tool targets Israeli entities and shares technical similarities with a backdoor previously linked to an Iranian threat actor.