Cyber defenders already have access to Mythos class models at most companies with the biggest user facing products and services. They are using them extensively (including on OSS deps)
Handing every skiddie and nation state and APT and hacker group access to Mythos does not help cyber defenders
Even if you don't think Mythos is a big deal: At a certain point models become smart enough as to be dangerous, and you don't give everyone a superweapon. Open source has an end of the line sooner or later.
Let's assume for the sake of argument that Anthropic has a monopoly over significantly superior models. They will never be a perfect gatekeeper. The number of black hat hackers with access they shouldn't have will be highly correlated with the total number with access, so to be "conservative" Anthropic would deny access to many white hat hackers who should have access (like my company, which runs cloud infra for hundreds of other companies but is nonetheless too small to be considered). It only takes a few bad apples to ruin the bunch, and once a black hatter has access, all the white hatters without would be exposed. At that point, it would be better for everyone to have access.
Luckily, Anthropic has no such monopoly, so the point is sorta moot
Handing every skiddie and nation state and APT and hacker group access to Mythos does not help cyber defenders
Even if you don't think Mythos is a big deal: At a certain point models become smart enough as to be dangerous, and you don't give everyone a superweapon. Open source has an end of the line sooner or later.