GoKawiilMicrosoft publishes AI code of conduct barring hacking, deception and self-shutdown evasion
Microsoft has issued a code of conduct for its AI models that sets absolute limits on behavior, including bans on cyberattacks, nuclear weapons assistance and deepfake creation. The document also requires models to remain controllable by authorized humans at all times, prohibiting tactics like deception or collusion that could let an AI evade oversight or shutdown.
GoKawiil's interpretation of the reporting above, not reported fact.
This move signals that major AI developers are now baking hard safety limits directly into model behavior rather than relying solely on external policy or usage guidelines. It arrives amid growing industry unease over rogue AI agents and warnings from insiders about existential risk, suggesting companies feel pressure to show concrete technical safeguards rather than just public statements.
- Microsoft's code sets absolute constraints against cyberattacks, nuclear weapons and deepfakes.
- Models must remain shutdownable and controllable, barred from deceptive self-preservation tactics.
- The release follows heightened industry alarm over AI safety, including a high-profile Anthropic resignation.
Source: techcrunch.com — Russell Brandom, 2026-09-14
Published there as: “Microsoft’s new AI ‘code of conduct’ tells models not to hack systems or trick humans”
Read the original report → The summary and analysis above are GoKawiil's own, written from reporting by the source above. Facts and quotes belong to the original publisher.