Guide to Trustworthy AI
The Art of Not Breaking the World with AI
Presentation Summary
In this high-energy session from the Executive Forum at Dynamics Minds, the Cloud Lighthouse team dives headfirst into the realities of building and breaking AI systems. With real-world demos, hard truths, and plenty of humour, they explore what it really takes to make AI trustworthy, from Microsoft’s six responsible AI principles to jailbreak prompts, red teaming, and grounding data strategies. You’ll learn:
- Why trustworthy AI starts with cultural change, not just compliance.
- What happens when AI goes rogue (with some eyebrow-raising live demos).
- How to use maturity assessments and defect rates to make AI safe for production.
- Why data governance, risk mitigation, and shared responsibility are non-negotiable.
Whether you're building your first AI solution or buying third-party tools, this session will challenge your assumptions and leave you with a roadmap and a few laughs.
Presentation
Ana Welch
Chris Huntingford
