How a session in Amsterdam is put together
Every keynote is tailored after a briefing call, and the larger engagements start before the day rather than on it: a diagnostic of the client's own people, so the room hears its own evidence rather than somebody else's. What the room leaves with is written down.
What that has looked like: about 1,500 people joining a session run hybrid from central London at short notice, for a technology business changing AI tools faster than its people could absorb, which became executive coaching for the transformation lead; and an agency business convening Asia, Asia-Pacific, EMEA and North America into one event, where the work was naming the recurring situations nobody had vocabulary for, and the company took that wording into its strategy.
Six of the seven engagements published on this site produced further work after the session rather than at it. The seven are set out in full at case studies, each with a line saying what it does not show.
How to read these examples. They are published to show how the work is put together, not as a client list for Amsterdam. Most engagements are covered by confidentiality and are described without naming anyone; where a client is named anywhere on this site, it is with their agreement.
A city that wrote automation bias into its own register
Amsterdam publishes a register of the algorithms it uses. The entry for its welfare investigation model, now marked buiten gebruik, out of use, contains this under the heading on human intervention: there is no automated decision-making, there is meaningful human involvement before a decision is taken, and work instructions are being drawn up om teveel vertrouwen in de uitkomst van het model (“automation bias”) te voorkomen, to prevent too much trust in the model’s outcome.
A municipal register, in Dutch, naming automation bias as the thing the work instructions exist to prevent. And the same entry records that the model tested fairer than the caseworkers on almost every sensitive characteristic. The city withdrew it anyway.
That is a harder and more interesting case than the usual one, because the easy story is that the algorithm was biased. Here it was less biased than the humans, and the city still decided that was not sufficient reason to keep it. Whatever you conclude from that, it is a real decision made by a real institution with the evidence in front of it.
And then said in public what human-in-the-loop becomes
On 4 August 2026 the Amsterdam AI Lab published work on human-in-the-loop as a design principle. Its conclusion, in the city’s own words: human-in-the-loop is a precondition that means something different to everyone, and in practice the easiest form usually wins, een vinkje op een lijst, a tick on a list. The risk, it says, is that human involvement is then formally arranged and adds little.
They built five playable prototypes to demonstrate it, including deliberate anti-examples. One is nicknamed the doorklikmachine, the click-through machine, in which the green button is larger than the red one and disagreeing with the system requires a lengthy justification.
A city government publishing a working demonstration of how its own safeguard fails. That is the argument at human in the loop is not a safeguard, made by somebody else, with a prototype attached.
The case everybody half-remembers, and the sentence they miss
The Dutch childcare benefits scandal is the most cited failure of automated government in Europe, and the detail that matters for this argument is almost always left out. The parliamentary inquiry’s final report, Ongekend onrecht, of 17 December 2020, records that de door het systeem geselecteerde aanvragen worden handmatig extra gecontroleerd: the applications selected by the system were additionally checked by hand.
There was a human in the loop. The government fell anyway. Anyone using that case to argue for human review is arguing for the thing that was already there.
The national picture is consistent with the municipal one. The Netherlands Court of Audit found 433 AI systems across 70 government organisations with only about five per cent in the public register, and wrote that there is an incentive to rate risks as low. Dutch enterprise AI use is 33.2 per cent on Eurostat’s 2025 figures, sixth in the Union, with a break in the series that year.
Who books this in Amsterdam, and who should not
Municipal and national government, banks and insurers, the technology and scale-up community, and the conference circuit around World Summit AI. Dutch audiences ask direct questions early and do not soften them, which suits a talk whose sources are published.
One honest correction to carry: the Dutch impact assessment for algorithms is recommended rather than mandatory, whatever a briefing note may say, and the binding obligation is the AI Act article it was rewritten to serve.
The three keynotes a Amsterdam room can book
Keynotes run 30 to 90 minutes, in person or virtual. A 30-minute main-stage version of each talk is available for conferences; the full argument needs 40 or more. The sector examples change for the room; the argument does not. All three, with the showreel, are at keynotes, and the programme copy is at the speaker pack.
Drift versus Design
Drift versus Design: why most organisations hand their judgement to AI without deciding to, and how to design your way through instead.
Most organisations are adopting AI by drift: a thousand reasonable decisions that add up to judgement nobody chose to give away. Rahim Hirji, author of SuperSkills, shows leaders which mode they are running, where AI sharpens judgement and where it weakens it, and hands them the controls.
We Are Superheroes
We Are Superheroes: AI is the suit. The human decides.
AI makes everyone faster, stronger and more capable. It does not decide; people do. Through a family story across four generations and three continents, Rahim Hirji hands the audience the seven human skills that grow more valuable as the tools spread. They arrive thinking AI is the story and leave knowing they are.
WTH (What the Human)
WTH (What the Human): a live test of a board's own judgement.
Business is being rewired as AI arrives, and human judgement is leaving with it. In three acts, Rahim Hirji shows a board what AI is changing in how organisations run, then ends with a live test of the board's own judgement, in the room and in real time. Nobody forgets the result.
Where these sessions happen in Amsterdam
- RAI Amsterdam
- 116,200 square metres hireable across twelve halls and seventy conference rooms, with the RAI Theatre at 1,750 seats. The city's default for anything at conference scale, and where most of its international events happen.
- Beurs van Berlage, city centre
- The Grote Zaal takes 1,600 and the Effectenbeurszaal 800, with 4,270 square metres across the building. The former stock exchange, and the right room when the audience should feel they are somewhere rather than in a hall.
- Westergas, the Gashouder
- 3,500 in a circular former gasholder. Industrial character, no fixed seating, and a production budget to match.
- Taets Art and Event Park, Zaandam
- Hall 1 alone is over 6,320 square metres. Just outside the city and the home of World Summit AI.
- Johan Cruijff ArenA
- Twenty-two conference rooms alongside the stadium, with dinners to 1,000. Useful when the brief wants scale and a name rather than a conference floor.
World Summit AI runs 7 and 8 October 2026 at Taets, its tenth edition, and is the obvious platform in this city for this argument. IBC returns to the RAI 11 to 14 September 2026 with 44,000-plus attendees, and Money20/20 Europe is an Amsterdam event at the RAI rather than a London one.
Three corrections for anyone building a Dutch list from older material. Integrated Systems Europe has left Amsterdam for Barcelona. Amsterdam Drone Week is discontinued, on its own domain’s statement. And TNW Conference has no announced future edition.
