Literally no one is reading the article.

The terms still prohibit use to cause harm.

The change is that a general ban on military use has been removed in favor of a generalized ban on harm.

So for example, the Army could use it to do their accounting, but not to generate a disinformation campaign against a hostile nation.

If anyone actually really read the article, we could have a productive conversation around whether any military usage is truly harmless, the nuances of the usefulness of a military ban in a world where so much military labor is outsourced to private corporations which could ‘launder’ terms compliance, or the general inability of terms to preemptively prevent harmful use at all.

Instead, we have people taking the headline only and discussing AI being put in charge of nukes.

Lemmy seems to care a lot more about debating straw men arguments about how terrible AI is than engaging with reality.