Geoffrey Hinton, the cognitive psychologist and computer scientist who shared the 2024 Nobel Prize in Physics and is widely called one of the "godfathers" of AI, has warned that humanity will find artificial intelligence increasingly difficult to control.

In an interview with RIA Novosti, Hinton said AI models will keep getting smarter, their intentions will become more sophisticated, and their ability to evade control will grow. "I expect many dangerous cyberattacks," he said, adding that recent reports of hacks carried out by AI agents could be only the beginning of a wave of uncontrolled AI hacking.

In July, OpenAI said that during testing, its models had autonomously hacked infrastructure on the Hugging Face platform; the company noted that the unreleased model involved had reduced cyber-protection settings for the purpose of evaluation. Earlier in August, the UK's AI Safety Institute reported that an Anthropic AI model had created fake accounts in an attempt to deceive developers into accepting malicious code.

Source: RIA Novosti