Microsoft (MSFT) has published a new AI "code of conduct", which aims to avoid the development of code that evades human oversight, The Information reports

Such announcements have historically carried little direct earnings relevance; where they have mattered is as signalling, both to regulators weighing intervention and to enterprise customers whose procurement decisions increasingly weigh vendor safety posture, a channel that operates through deal flow rather than through any visible line item. The detail that the code targets development evading human oversight aligns it with the internal-controls language regulators have themselves been gravitating toward, which suggests positioning rather than concession. Two actors matter beyond the company: the regulator, whose response will indicate whether self-governance is being accepted as sufficient, and the peer set, since comparable frameworks from rival firms have tended to follow quickly once one major player commits publicly. The follow-ons worth tracking are whether the code is enforced with named accountability, whether peers match it, and whether it surfaces in upcoming antitrust or AI-safety proceedings, where such documents have previously been cited as evidence of good faith.