Comment by lostdog

Those behaviors have been anticipated for years if not decades. And each incident is minor and leads to clearer rules and safety behaviors for AI agents.