In the next five years, artificial intelligence (AI) systems may appear that surpass humans in all cognitive tasks, creating a risk of losing meaningful human control over them. This warning was issued on September 23 by the Alan Turing Institute in the UK.
“Questions about the long-term future of humanity deserve attention,” said George Williamson, director general of the institute. “However, discussions about advanced artificial intelligence should not be conducted exclusively by technology companies or built only around the most extreme scenarios. We know enough about the many serious risks associated with AI to act.”
The report noted that within five years, more powerful and autonomous AI systems could emerge, whose behavior would become increasingly difficult for people to evaluate, control, or cancel if necessary. The institute emphasized that uncertainty regarding long-term threats should not hinder efforts to address existing risks.
Additionally, the study highlighted threats to cybersecurity and democratic processes, the potential for AI actions to diverge from human intentions, as well as chemical and biological threats. The institute called for strengthening verification and control mechanisms for both individual AI models and full-fledged systems after implementation.
Separately, Sam Altman, CEO of OpenAI, warned on September 14 that AI progress could follow two extremely dangerous scenarios that humanity must avoid: first, the potential risk of a complete loss of control over future development in favor of AI systems; second, an excessive concentration of unlimited power in one individual’s hands.
