Anthropic released its Mythos AI model to about 40 select companies including Apple, Microsoft, and Google, while withholding it from the public citing cybersecurity risks. The nation's top cyber defense agency (CISA) was excluded from access. The post argues this selective strategy amounts to a corporate cartel that sets a dangerous precedent—but cybersecurity experts are split, with some calling it responsible disclosure and others criticizing it as market gatekeeping designed to entrench incumbent vendors . Days after the announcement, unauthorized users were reported to have gained access through a third-party vendor anyway.