Trusted Agentic AI
for Decision Makers
Four sessions. Small groups. Technically grounded, strategically focused.
Learn to ask the right questions, anticipate the hard ones, and align your stakeholders.
Enroll nowFour Sessions. Six Hours. One Framework.
Every Wednesday, 14:00 CET / 8:00 AM ET. Maximum 6 participants for meaningful interaction. Chatham House Rule throughout.
Every participant starts with a 30-minute intake call to discuss your challenges and tailor the sessions.
All recordings, reference materials, and assessment tools remain available after the programme.
What We'll Cover
We go deep enough to challenge your technical teams, but you don't need to be one of them. Pre-reads bring everyone up to speed, and each session focuses on the thinking, not the code.
P Session 1 Potential
- Where we actually are — what's real, what's overblown, and the gap between impressive demo and reliable production
- From model to agent — what makes an agent different from a chatbot, and why getting the right information to it at the right time is the key challenge
- Adoption and autonomy — traditional software passes intent, agents create it. Where that shift changes the risk, and where it creates value
- The landscape — frontier vs. open-weight models, cloud vs. private/on-prem, what you're locked into, and where your data actually goes
- Protocols and standards — the emerging standards that matter for long-term investment decisions
- How do you know it works? — measuring agent performance, the #1 gap between pilot and production
- Leave with — the right questions to ask your business, product, and AI engineering teams
A Session 2 Accountability
- Shadow agents — agents already running in your organisation without governance oversight, and how to spot them
- The liability chain — when an agent makes a bad decision, who's responsible? We walk through a real scenario
- Regulatory landscape — EU, US, and UK approaches, where they converge, and what auditors actually look for
- Audit trails and evidence — what you need to log and why, designed for compliance not just debugging
- Data governance and privacy — what data agents can access, where it goes, and what regulations require
- Mapping your exposure — learn to recognise where the gaps are before your auditor or regulator does
- Leave with — the right questions to ask your legal, compliance, and risk teams
C Session 3 Control
- Policy vs. architecture — policy says "don't," architecture says "can't." Why the difference matters for agents
- Identity for agents — who is this agent, who does it act for, and how do you prove it? The landscape of solutions emerging now
- Authority and delegation — what an agent can access, for how long, and what happens when agents delegate to other agents
- Zero trust for agents — containing what agents can do, not just what they should do, including across organisations
- What exists today — what tooling is available now, what's emerging, and where the gaps are
- Putting it together — we walk through a real-world scenario and show what a control architecture looks like in practice
- Leave with — the right questions to ask your security, identity, and platform teams
Session 4 Your Roadmap
- Bring your challenge — three weeks of questions, conversations, and observations. Now bring what surfaced
- Embracing the wicked problem — competing priorities, alignment on trade-offs, iterative over waterfall
- Pressure-test your roadmap — challenge each other's assumptions, cross-pollinate solutions across participants
- Leave with — a roadmap you can defend across all stakeholders, and the conversations to have this month
Questions you'll be equipped to answer
Shane Deconinck
Trusted AI Agents · Decentralized Trust
I've been writing about the trade-offs at the centre of agentic AI: competitive, responsible, and aligned with regulation, all at once. Most guidance picks one angle. This programme looks at all three, so you can anticipate the questions from every stakeholder and align them. Pragmatic frameworks, protocol deep-dives, and trust architecture, read by professionals across the industry.
This space moves weekly. Standards shift, new protocols land, regulations evolve. A textbook is outdated before it prints. Every cohort gets material that reflects what's happening right now, not what was true six months ago.
- Invited panelist at the EU Parliament on AI governance
- 2025 LF Decentralized Trust Community Award
- Co-founder, LF Decentralized Trust European Chapter
- Speaker at Linux Foundation summits
What You Get
You'll leave with a governance roadmap grounded in your organisation's actual context, the right questions to ask your technical teams, and the vocabulary to bridge engineering, compliance, and leadership.
- 30-minute prep call before the programme to discuss your challenges and tailor the sessions
- 4 live sessions, 90 minutes each, maximum 6 participants
- Challenge submission: bring your organisation's real problems and get them worked through the PAC framework
- Chatham House Rule: speak freely about real challenges without attribution
- Recordings of instructor-led content and framework walkthroughs (group discussions are not recorded)
- Pre-session profiling: Use the Agent Profiler to map your use cases before we start. Your profile feeds directly into the sessions
- Reference materials: the PAC framework, protocol explainers, and assessment tools
- Completion record from trustedagentic.ai
This is for you if:
- You're responsible for AI strategy, governance, or architecture in an enterprise exploring agentic AI.
- You need to advise leadership on agent deployment but lack a structured framework.
- You're caught between teams moving fast and teams saying stop, and need to bridge both.
- You have a real challenge you want worked through. Not just theory, your actual situation.
This is probably not for you if:
- You want hands-on coding labs. The programme covers the technical mechanics, but you won't be writing code.
- You're looking for a beginner introduction to AI. This programme goes beyond the surface.
- You want vendor-specific guidance. This is deliberately neutral and standards-based.
The framework is jurisdiction-agnostic. Sessions reference the EU AI Act, NIST, and ISO 42001. Participants from any region apply it to their own regulatory context.
Join the April Programme
Live online via Zoom · April 8 – 29, 2026
Wednesdays 14:00 CET / 8:00 AM ET · Maximum 6 participants
€4,500 excl. VAT · approx. $5,300
Includes prep call, four live sessions, recordings, and all materials.
Questions or looking for a private session? Get in touch.
Looking for something more tailored? Explore consultancy →
Not ready yet? Get notified about future cohorts and new programmes.