Tech companies are increasingly warning that their own artificial intelligence creations could pose existential risks to humanity, but critics argue this apocalyptic narrative serves to distract from current harms and inflate stock prices. Anthropic, the AI firm behind the Claude Mythos model, recently claimed the system's ability to find cybersecurity bugs surpasses human experts and could have severe consequences if misused. However, some security experts doubt these claims.
This pattern is not new. Executives at leading AI providers regularly issue dire warnings about their products. In 2019, OpenAI delayed the release of GPT-2 over 'concerns about malicious applications,' only to release it months later. OpenAI CEO Sam Altman has previously stated that AI will 'probably most likely lead to the end of the world,' while also acknowledging that fears about GPT-2 were 'misplaced.'
Critics argue that focusing on hypothetical future catastrophes diverts attention from real-world issues such as bias, job displacement, and privacy violations. Shannon Vallor, a professor of data and AI ethics at the University of Edinburgh, says portraying AI as 'supernatural in its danger' makes the public feel powerless and suggests that only the companies themselves can be trusted to manage the risks.
In 2023, hundreds of tech leaders, including Altman, Anthropic's Dario Amodei, Bill Gates, and Demis Hassabis, signed a statement equating AI extinction risk with pandemics and nuclear war. The same year, Elon Musk signed a letter calling for a six-month pause on advanced AI development, only to launch his own AI company, xAI, less than six months later. Emily M Bender, a professor at the University of Washington, describes this as 'a pattern of unsubstantiated claims of power.'
