The biggest AI risk in the coming year isn't a neural-network takeover of the internet, but the use of AI as a weapon in information and cyber wars, said Vyacheslav Beresnev, executive director of the Association of AI Development Laboratories and an expert with the NTI NeuroNet market.

Speaking to RT, Beresnev called panic about a machine rebellion a media headline rather than a technical forecast. Real damage - potentially hundreds of billions of rubles or dollars - could come from targeted attacks on infrastructure or large-scale disinformation campaigns, not from losing control of algorithms, he said.

He argued that calls for urgent regulation often sound noble but in practice shut out small companies and startups while entrenching the dominance of giants, and that one-sided slowdowns create dangerous geopolitical asymmetry.

Instead of trying to artificially halt progress, Beresnev proposed that the global community develop transparent "traffic rules" similar to those governing nuclear energy. In this new reality, he concluded, the winner isn't whoever has the more powerful AI, but whoever learns to manage it safely and effectively.

Earlier, US President Donald Trump rejected claims that AI could take over the world and destroy humanity, calling such talk a deception and saying he wasn't worried about AI slipping out of control.

Source: Lenta.RU