Microsoft’s new AI ‘code of conduct’ tells models not to hack systems or trick humans
techcrunch.comRussell Brandom
Microsoft has published a code of conduct establishing principles and safety constraints for its AI models, including directives against hacking systems or deceiving humans, while emphasizing support for human agency rather than replacement.
Why it mattersAI Tools Daily readers building with Microsoft's models need to understand these official behavioral expectations, as they shape what capabilities and responses to expect from the tools they integrate.
From the source: The code of conduct lays out general principles that Microsoft AI models should uphold — supporting humans rather than replacing them, for instance, and accelerating human flourishing — as well as specific safety constraints meant to implement those principles.Read at techcrunch.com