Tech · 19 d ago
Toronto researcher warns AI worms make poorly secured smart devices easy targets
A computer worm
self-spreading malware
In June, a team led by University of Toronto professor Nicolas Papernot showed AI can power self-spreading malware that adapts its attack to each device it finds, from laptops to cameras.
This week Anthropic, an AI company, said small AI models may be "months away" from copying themselves onto devices such as smart ovens and off-brand Android TV boxes.
What we know
- Hundreds of French supermarkets hit
- Controllers made by Danfoss targeted
- Losses estimated in the millions
Not yet
- Any role for AI
Overnight Sunday, hackers took control of refrigeration systems at hundreds of supermarkets in France, forcing fridges into defrost mode. Nothing reported so far suggests AI was involved.
More than 400 scientists, including Geoffrey Hinton and Yoshua Bengio, signed an open letter on Tuesday asking that AI capable of cyberattacks be tested cut off from the internet, and for an international watchdog.
Protect your devices
Papernot's advice: stop reusing the same password for several services, and keep every device up to date, because the time between a fix and an attack "is becoming very, very short".
