Such stances are very easy to make when there is no financial incentive or no competition with any meaningful money on the line. But as soon as there is competition, they don't care about "AI policies" if it is the means to make a lot of money.

The best example of this is security in open source software which maintainers have no choice but to use AI to protect themselves before an attacker can use AI to find and build exploits for security vulnerabilities faster than they can patch it.

In some sectors such as banking, crypto, defense and government, this is extremely critical. It will soon be the same with AI companies who have the AI model and someone could use AI to breach them with the goal of leaking the proprietary model.

So to them, policies like this, are actually a downgrade that they cannot afford to do.