Your AI is only as responsible as you are
Recorded at Microsoft Build
Ryan welcomes Sarah Bird, Microsoft’s Chief Product Officer for Responsible AI, to discuss how we can build and use AI responsibly with the NIST approach, why most irresponsible AI comes from experimentation without thought of impact, and how Microsoft is researching thoughtful human/AI workflow design to reduce unnecessary escalation.
The NIST Approach to Responsible AI
Sarah Bird explains that the NIST AI Risk Management Framework provides a structured methodology for organizations to assess and mitigate risks throughout the AI lifecycle. Key principles include:
- Govern – Establish organizational structures and policies for AI oversight
- Map – Understand the context, risks, and potential impacts of AI systems
- Measure – Implement quantitative and qualitative metrics to evaluate AI behavior
- Manage – Continuously monitor and respond to identified risks
Why Irresponsible AI Happens
According to Bird, most irresponsible AI deployments stem from experimentation conducted without adequate consideration of downstream effects. Teams often focus on technical capability first, treating safety and responsibility as afterthoughts rather than integral design requirements.
Human/AI Workflow Design Research
Microsoft is actively researching how to design human/AI workflows that reduce unnecessary escalation. This involves:
- Building friction points that prompt user reflection before high-stakes actions
- Designing interfaces that make AI limitations transparent
- Creating feedback loops that help users calibrate their trust appropriately
Connect
This episode was recorded at Microsoft Build. Listen to our other episode recorded at Build on agentic workflows here. Connect with Sarah on LinkedIn.
Comments
No comments yet. Start the discussion.