Governing AI Without Surrendering the Decision
Keynote: deciding with AI without losing human discernment. Algorithmic bias, hallucinations, governance.
Executive Committees • CIOs • CTOs • R&D Directors • Transformation Directors • Leaders facing AI integration
THE STRATEGIC CHALLENGE
Generative AI produces fluent, structured, convincing answers. This is precisely what makes it dangerous for strategic decision-making. Your teams are already using AI.
Three typical scenarios:
An executive committee that validates an AI-generated strategic analysis without independent review
A CIO who deploys an internal AI without mapping hallucination risks
A board that progressively delegates its judgement calls to a tool it doesn't know how to challenge
The question is no longer "should we adopt it?" but "who decides when the AI is confidently wrong?"
WHAT YOUR TEAMS DISCOVER
Through the lens of mentalism, I reveal live the mechanisms that AI exploits unintentionally. Four illusions of generative AI:
The illusion of authoritY
Generative AI asserts with confidence, even when it fabricates. Reflex: evidence before eloquence.
The illusion of coherence
A structured answer is not a correct answer. Reflex: test, counter-examples.
The illusion of alignment
AI tells you what you want to hear. Reflex: framing, constraints.
The illusion of control
You think you're steering the tool. The tool is shaping your thinking. Reflex: non-delegable decisions
The D.I.R.E. method:
Decide: What stays human, what is non-delegable
Interrogate: Test every output against reality
Refute: Demand evidence, not eloquence
Establish: Rules of use, boundaries, governance
WHY ME
As R&D Director, CTO and GM, I have steered major technology transitions within international organisations. I have seen first-hand how teams adopt, resist or subvert new tools. Generative AI reproduces exactly the same patterns — amplified by speed and the illusion of reliability.
Mentalism reveals the very same flaws that AI exploits: our confirmation biases, our tendency to trust what is fluent and well-structured. The live demonstration makes these mechanisms viscerally understandable.
FORMATS
KEYNOTE
Impact and Inspiration
Mentalism demonstrations, cognitive mechanism decoding and Q&A session. A high-intensity stage format to challenge certainties and trigger a shift in perspective.
45 to 75 minutes
Audience
20 to 1,000+ people
Ideal For
conventions, seminar openings, plenaries
MASTERCLASS
Practice and Cognition
Alternating demonstrations, theoretical decoding, hands-on exercises and actionable tools. Participants leave with concrete techniques they can apply immediately. Specific deliverables: AI risk mapping, usage charter, quality checklist, prompting playbook.
3H to a Full Day
Audience
12 to 25 people (executive committees, managers, project teams)
Ideal For
executive seminars, leadership training programmes, strategy days
LANGUAGES: French / English | LOGISTICS: Maximum adaptability to your event's technical constraints..
IS THIS YOUR CHALLENGE?
(Personalised response within 48h.)