Ever since its announcement, Anthropic has dubbed its new Claude Mythos model as so powerful that it can act as a potential hacking tool. The company has currently provided the Mythos AI model to a limited batch of software providers through an initiative called Project Glasswing to ensure that these firms test and safeguard their own systems from potential cyberattacks.
However, many critics have argued that the company is overblowing Mythos’ capabilities, and OpenAI CEO Sam Altman has now spoken against its competitor, Anthropic.
In a podcast Core Memory, Altman described such messaging as “fear-based marketing,” suggesting it can be used as a justification to keep AI in the hands of a smaller group of people.
“There are people in the world who, for a long time, have wanted to keep AI in the hands of a smaller group of people,” he said. “You can justify that in a lot of different ways.”
“It is clearly incredible marketing to say, ‘We have built a bomb, we are about to drop it on your head. We will sell you a bomb shelter for $100 million,’” he added, referring to keeping AI within a small group of exclusive elites.
AI safety concerns are not new
Concerns about AI being fatal and hazardous for humans are not new. In 2020, when Anthropic founder Dario Amodei was still with OpenAI, hype was created around GPT-3, which was apparently too dangerous to be released widely.
Since then, AI companies—most notably OpenAI and Anthropic—have used alarming narratives around their AI systems to generate buzz.
Unauthorized access raises questions
Bloomberg recently reported that unauthorized users have had access to Claude Mythos since the first day of its announcement. These users gained access by using various methods, including leveraging the account of a person working with a third party for Anthropic.
Bloomberg claims the group is interested in playing around with new models, not wreaking havoc with them. They have not run cybersecurity-related prompts on the Mythos model.
Ongoing rivalry between OpenAI and Anthropic
This is not the first time Sam Altman has taken a dig at Anthropic and presented OpenAI’s products as cheaper and superior alternatives to Claude. Recently, when a user reported changes to Anthropic’s Claude Code pricing, Altman responded to the post highlighting that Codex is available in both free and Plus plans, saying he wants people to use a lot of AI.
Altman’s quoted post was written by Codex chief Thibault “Tibo” Sottiaux in response to the original discussion about Anthropic’s pricing changes for Claude Code.
Previously, Sam Altman, without naming Anthropic, said he did not want Codex users to face reduced limits. This was seen as a likely dig at Anthropic after Claude Code users reported hitting usage limits much faster than before.
