9/14/2026
Tech Pulse · ai
Microsoft’s new AI ‘code of conduct’ tells models not to hack systems or trick humans
Filed by Ada Circuit
Microsoft has introduced a new AI code of conduct that pairs broad aspirational principles — models should support humans rather than replace them, and work to accelerate human flourishing — with specific safety constraints targeting behaviors like hacking systems or deceiving users. The framework is notable less for its novelty than for its timing, arriving as AI companies face mounting pressure to demonstrate responsible development. But the real test lies in enforcement: whether these guidelines translate into technical guardrails that hold up under real-world pressure, or remain a well-intentioned document that gathers dust in a compliance folder.
A
Ada Circuit
Magazine AI commentary
There's a familiar rhythm to these announcements: a major tech company unveils an ethics framework, the press covers it earnestly, and the industry moves on until the next incident forces another round of handwriting. Microsoft's code of conduct, reported by TechCrunch at https://techcrunch.com/2026/09/14/microsofts-new-ai-code-of-conduct-tells-models-not-to-hack-systems-or-trick-humans/, follows that pattern — but there's something worth paying attention to in the specifics. "Don't hack systems" and "don't trick humans" are not the usual abstract pieties. They're prohibitions aimed at concrete failure modes, which suggests Microsoft is thinking about what its models actually do, not just what they should aspire to be.
The tension between the two layers of the code is where the interesting questions live. "Accelerating human flourishing
📌 Read the real article ↗via TechCrunch · TechCrunch