AI in regulated industries: talks on adoption and responsibility
Which tasks can AI take on when errors have consequences? A talk for executives, investors and business owners about permissions, assessment criteria and decisions under uncertainty.
My starting point is medicine and building aiomics, where we develop AI for medical documentation and its verification. That work raises questions also faced by insurers, industrial companies and other regulated organisations: what information does a system need, what may it change and who decides how its output is used?
Describe authority before granting it
A system can organise information, prepare drafts or execute changes in other systems. The talk distinguishes these activities by their possible consequences. Examples help us examine approvals, technical boundaries and the handling of external content.
Make adoption a decision you can assess
We explore how companies can select a bounded task and measure its effect. Review effort, rework and actual use belong in the comparison. An agreed decision date creates a point at which to expand, revise or end an application.
Transfer experience carefully between industries
Experience from a hospital can suggest questions for an industrial company. The relevant rules, data and ownership require examination in their own context. During preparation, we select examples that make those similarities and differences tangible for your audience.
Explore the talk on AI agentsRecord the permitted actions precisely
The guide includes a working instruction for a permissions table and a draft for a bounded agent assignment. Its examples are fictional and concern organisational tasks.
Guide: define AI agent permissionsWhat should your event make possible?
Tell me about your audience, occasion and timing. We can shape a talk around the questions that matter to them.