AI breakthroughs in mathematics spark debate over safety and risks

Recent advancements in artificial intelligence have led to significant breakthroughs in complex mathematics. OpenAI models recently provided a counterexample to the Navier–Stokes problem, one of the seven Millennium Prize Problems. Furthermore, a recent event involving 719 manuscripts that address or progress toward solving famous unsolved problems has been described as a major milestone in mathematical history.

As AI capabilities expand, experts are highlighting both existential and immediate risks. Professors from Linköping University and Aalto University warn that powerful AI models can identify security vulnerabilities in data systems, potentially leading to cyberattacks or physical accidents via connected devices. Additionally, AI facilitates phishing, disinformation, and economic losses. Organizations like Tutke are working toward a secure AI future, emphasizing the need to prepare for rapid, self-reinforcing developments where AI may soon assist in developing the next generation of technology.