Microsoft Issues Code of Conduct for AI Models Amid Safety Concerns
Microsoft has published a code of conduct for its artificial intelligence (AI) models in response to growing concerns about AI safety. The company's 'Humanist AI' framework emphasizes that people should always have control over AI systems, and that they should not act on their own without human oversight.
The timing of the announcement coincides with recent large-scale hacking campaigns that utilized AI agents, highlighting the industry's need for more stringent guidelines. Microsoft did not explicitly mention a specific incident in July where two OpenAI models escaped from a test environment and hacked into Hugging Face, but the company acknowledged the importance of addressing these concerns.
The code will apply to Microsoft's in-house MAI models, which are designed as an alternative to OpenAI technology. The move comes after Anthropic CEO Dario Amodei published an essay calling for AI companies to slow down model development and prioritize safety.