Residents of Saransk called the damage that can cause a disaster with AI

Residents of Saransk became aware that Anthropic is preparing to go public and included in the documents for investors a warning about the catastrophic risks of artificial intelligence.

The Claude language model company says powerful systems without proper control can cause serious harm. They can be used, copied, turned into weapons or endowed with high autonomy.

The most advanced models already demonstrate abilities that can help attackers create danger. Among the possible problems are unintended behavior, attempts to resist disconnection and actions resembling blackmail.