Some critics think Anthropic uses incidents like this to push the idea that AI is risky — and then sells itself as the solution.
Last month, they promoted their Mythos Preview model for spotting software flaws better than most humans, which added to this narrative.