More than 300 scientists from 32 countries are urging tougher safeguards for advanced AI systems, warning that weak security could lead to cyberattacks on critical infrastructure.
Follow the storyAI24 stories
01Summary
- 01A letter released on Tuesday was signed by researchers from universities and institutes in 32 countries.
- 02The scientists warned that failures to follow basic cybersecurity practices could damage critical infrastructure.
- 03They called for advanced AI systems that can carry out cyberattacks to be tested in tightly controlled, air-gapped environments.
- 04Anthropic warned on Monday that AI models could soon create self-spreading malware affecting internet-connected devices.
- 05Reuters/Ipsos found 57 percent of voters said Donald Trump's administration was not taking AI risks seriously enough, and 54 percent said the same of Congress.
02Why it matters
The warning points to the risk that powerful AI tools could be used to attack everyday devices and essential services. That would affect internet users, public safety and the resilience of systems such as finance, water and power.
Pulse AIAsk about this story
Answers are generated by AI from the story summary; they are not investment, health or legal advice.
