Home / Technology / AI Giant Accused of Hype Over Cybersecurity Threat
AI Giant Accused of Hype Over Cybersecurity Threat
12 Apr
Summary
- AI firm Anthropic withheld a powerful model, citing cybersecurity concerns.
- Skeptics question if this is hype to attract investment rather than real risk.
- The company has secured significant media attention for its AI developments.
Anthropic has announced it is withholding a powerful AI model named Mythos, citing overwhelming cybersecurity concerns. This announcement has drawn attention from government officials and prompted a letter to the UK government highlighting potential catastrophic cybersecurity risks.
However, prominent figures in the AI community express skepticism, suggesting Anthropic may be employing hype and exaggeration to attract investment, similar to tactics used by rival OpenAI. The company has recently enjoyed extensive media coverage, including features in The New Yorker, Wall Street Journal, and Time magazine.
Experts question the substantiation of Mythos's capacities, with some suggesting vague marketing language obscures evidence and raises concerns about seeking investment without scrutiny. The decision not to release Mythos prevents independent evaluation of the company's claims.
Further complicating matters, Anthropic has faced challenges with computing capacity for its existing popular models, including usage caps and requiring extra purchases for third-party tools. This infrastructure limitation may also contribute to the decision not to release a new, highly anticipated model.
Critics argue that Anthropic's strategy of emphasizing safety publicly, while potentially prioritizing profits, mirrors a 'bait and switch' playbook observed with OpenAI. The company's publicity efforts have been noted for effectively obscuring this potential shift.
