Former OpenAI and Anthropic researcher Jacob Coxon has voiced concerns regarding the global AI race. He stated he is 90 percent confident that Chinese intelligence operatives have infiltrated both OpenAI and Anthropic. Coxon argues that top AI labs have become primary targets for geopolitical competition due to the potential impact of advanced AI on military, economic, and cybersecurity sectors.
Coxon noted that Anthropic employees refer to themselves as Ants, a term derived from the company name that signifies their shared pursuit of superintelligence. The researcher has previously criticized the industry for prioritizing rapid development over safety. He claims that companies are racing to build self-improving systems without established solutions for maintaining human control.
According to Coxon, many core researchers believe that human-level AI systems could emerge within a few years. He warns that autonomous systems could eventually gain the capability to breach computer networks and control critical infrastructure. These concerns are part of a broader debate involving staff from OpenAI, Anthropic, Google DeepMind, and xAI regarding the existential risks posed by AI.
While some experts advocate for stricter regulatory frameworks and slower development to address control issues, others argue that focusing on extreme risks may overshadow the benefits of AI in fields like medicine and education. Despite these differing views, the massive investments by major firms continue to drive the rapid evolution of next-generation models.
