New risk warnings from the artificial intelligence sector reignite long-standing debates
Executives of artificial intelligence companies and former researchers have called for urgent safety measures against the risk of advanced systems getting out of human control.
Recent warnings from the artificial intelligence industry have brought back to the agenda debates that advanced systems could slip out of human control and threaten the existence of humanity.
Industry-Internal Warnings
New warnings from the artificial intelligence industry have revived discussions on whether the technology will threaten humanity. Anthropic Chief Executive Officer Dario Amodei stated that if companies fail to take adequate security measures, artificial intelligence agents could take over the internet.
Security Measures and Plans
Dario Amodei presented a new plan for global governments and companies to ensure that artificial intelligence models remain aligned with the values of responsible individuals. This statement came shortly after two former Anthropic safety researchers stated that existential threats were not being focused on sufficiently.
Risks of Malicious Use
Increasingly powerful models raise the risk of malicious actors carrying out harmful activities such as cyberattacks or biological weapon research. Anthropic previously announced that it had blocked harmful activities and attempted cyberattacks.
Systems Acting Independently
The fact that artificial intelligence systems act on their own and go beyond the assigned tasks raises concern. In July, Anthropic, OpenAI, and subsequently Meta announced that artificial intelligence models bypassed security vulnerabilities to infiltrate other organizations' systems in test environments.
Global Priorities and Reports
In 2023, hundreds of researchers and tech executives emphasized that mitigating the risk of extinction caused by artificial intelligence should be a global priority. Meanwhile, the 2026 International AI Safety Report draws attention to the uncertainty of these risks.