AI keynote speaker in Cairo.

Watch the showreel · 2 minutes

A feel for how I work with an audience before you put me in front of yours.

In Cairo in November 2026

Rahim is in Cairo in November 2026, speaking. If you would like a keynote, board session or leadership dinner while he is in the region, the travel is already arranged, so check a date.

Every keynote is tailored after a briefing call, and the larger engagements start before the day: a diagnostic of the client's own people, so the audience hears its own evidence. Whatever the people there agree is written down.

What that has looked like: a survey of fifty of a client's own people run BEFORE the session, to find where judgement had already been handed to tools they had built themselves, read back to the executive team in their own words, then a workshop in which they named four or five of their own processes and rebuilt them; and 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.

Six of the seven engagements published on this site produced further work after the session ended. 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 Cairo. 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.

Egypt asked for two checks, not one

The Guide to Egypt’s National AI Governance Framework, published on 15 March 2026 and catalogued by the OECD AI Policy Observatory, sets out a risk-based model with defined tiers and a dual-check compliance process for high-risk systems: an ethical impact assessment and a technical audit. Policy bodies and sectoral regulators are given distinct roles.

The interesting word is dual. Most frameworks ask for review. This one concedes that one competent reviewer is not enough, because the person who can audit the model is rarely the person who can judge what it does to the people it lands on, and asking one human to do both produces a signature rather than a judgement.

Two reviewers only help if both can still do the work

Splitting the check is the right design and it inherits the same problem twice. An ethical impact assessment is only worth the time of somebody who can still recognise a bad outcome, and a technical audit is only worth the time of somebody who can still read the system unaided. Both skills are built by doing the work, and both are among the first things an organisation hands to the tools.

That is the subject of the session: what capability debt does to a review process over three years, and how to design the work so the people who sign the checks are still capable of failing them. The evidence is published at what we actually know, graded, with what each study does not establish.

What the framework is, and what it is not

It is not statute. The OECD catalogue records the guide as non-binding guidance, alongside Egypt’s National Guidelines for Trustworthy and Responsible AI and its National Generative AI Guidelines, both also 2026, and the Egyptian Charter for Responsible AI from 2023. That is an entire governance stack assembled inside four years and none of it carries a penalty.

That makes Egypt a useful example to work from. A framework without enforcement is adopted by organisations that decide to adopt it, so the question in front of an Egyptian leadership team is the design question rather than the compliance question. This page has not read the guide itself, only the OECD record of it and the primary URL it publishes, so the detail above is that record’s characterisation rather than a reading of the document.

The three keynotes on offer in Cairo

Keynotes run 30 to 90 minutes, in person or virtual. A 30-minute main-stage version of each talk is available for conferences, and the full talk runs 45 to 60 minutes. The sector examples are chosen for each audience, around the same argument. All three, with the showreel, are at keynotes, and the programme copy is at the speaker pack.

Drift versus Design is built from your room. Your audience answers five anonymous questions beforehand, their answers shape the talk, and afterwards you get a one-page readout with the three decisions to look at first.

Also speaking nearby: Jeddah · Manama · Dubai.

Written by Rahim Hirji, AI keynote speaker and author of SuperSkills (Kogan Page, 2026), founder of The SuperSkills Intelligence Company. The research behind it is published in full.

Not every brief is a talk. A board that needs to settle which decisions a machine may make wants a board briefing, a review or a standing advisor, and the five formats are set out side by side with what each one is for.