Experts warn of growing AI risks, call for guardrails
As artificial intelligence (AI) advances, researchers are expressing growing concern. Two former Anthropic employees issued public statements warning that the human race could be approaching extinction. Moreover, Anthropic CEO Dario Amodei warned about potential risks from the rapid development of AI, calling for it to «slow down,» the BBC reported.
In his latest essay, he raised two main concerns: AI’s ability to self-improve and even build the next generation of AI, and the difficulty of controlling these systems if they do things they weren’t asked to do. In addition to the risk of losing control of AI systems, Amodei also cited risks related to AI-enabled cyberattacks, bioterrorism and economic disruption.
Read also: The dark side of AI: Assessing the top cyber threats to Kazakhstan
The key measure Amodei proposes to navigate AI dangers without constraining technological advancement is to slow the pace at which AI models’ capabilities improve. «Progress will still seem fast, and we must make wise use of the time we gain,» he wrote.
Interestingly, the chiefs of two major competitors of Anthropic — xAI and OpenAI — Elon Musk and Sam Altman, respectively, both supported Amodei’s idea of slowing AI progress.
However, at the political level, AI is widely seen as a powerful source of leverage, pushing countries to develop more advanced tools. Commenting on concerns about AI’s growing capabilities, U.S. President Donald Trump dismissed alarming warnings as scenarios that would never happen.
Slowing AI progress is also difficult because of the rivalry between the U.S. and China in the field. Trump has already made it clear that he wants the U.S. to remain the industry leader because «whoever wins AI wins.»
Meanwhile, China is changing its national education programs, eliminating those it considers obsolete to make more room for AI and robotics. However, in 2025, President Xi Jinping proposed creating a global organization to oversee and govern AI and promote international cooperation. The U.S. rejected the idea at the time.
