Claims deserve evidence you can examine.
We examine behavioural risks, agent task claims and memory and user control in a clearly scoped review of systems that interact with people.
What we can work towards
- Scope, risk criteria and recorded test conditions
- Reproducible scenarios and finding records
- Prioritized recommendations for changes
- A plan for testing the changes again
Where to begin
Misleading trust, excessive attachment and boundary communication
Claims of sending or completing tasks compared with action records
Memory correction, deletion and current instruction handling
Let’s discuss your idea.
Start with a short note about your needs. Together, we can define a useful first conversation.
