Strategic foresight and signal intelligence
See what is converging on your institution, across capital, technology, regulation, and competition, early enough to act rather than react. Foresight built as a live capability, not a shelved report.
In plain terms: I watch what is actually moving across capital, law, and technology in your markets, test each signal before it reaches you, and wire the result into how you decide. Not a trends deck. A live read you can act on.
The practice is grounded in Operations Research and Systems Analysis (ORSA) discipline from US defense work, now run on AI agent stacks: scenario construction, assumption-based planning, and adversarial review executed by multi-agent systems and scored over time, so the foresight is auditable rather than rhetorical.
What a verified signal looks like
Three from the public record, each one dated and traceable, drawn from my Sovereign AI Landscape 2026 and the monthly tracker:
- Enforcement is real, not theoretical. Saudi Arabia's data-protection committees issued 48 decisions confirming violations with penalties during 2025 (Saudi Press Agency, January 2026). A regulator with a published enforcement record changes what "compliant" has to mean for anyone deploying AI there.
- Investment vehicles are being legislated, not just announced. Oman's Royal Decree 50/2026 (April 2026) created the region's first legislated AI Special Zone, and its data-protection law reached full enforcement in February 2026. A refreshed national programme and a purpose-built legal zone in one cycle is a signal, not noise.
- Skills commitments now carry numbers and dates. Lebanon signed an agreement with Oracle in December 2025 to train 50,000 people in cloud and AI over five years, with the minister stating up front that it involves no data exchange with the public sector. That is a deliverable you can hold to a date.
Every entry above survived adversarial review against its cited source before publication. That is the standard a signal has to meet before I bring it to you.
Who it is for
Governments, boards, and institutions that cannot afford to be caught flat by a shift they could have seen.
What you get
- Continuous signal intelligence across the fronts that matter to you
- Scenario and implication analysis tied to decisions
- A decision rhythm: scan, interpret, decide, score
- A foresight capability built into how your institution decides
How it works
- Stand up continuous, all-facet scanning
- Interpret signals into implications for you
- Wire foresight into real decisions
- Score the calls and calibrate over time
Questions
Is foresight the same as prediction?
How is this different from a one-off trends report?
Common questions
How much of this AI wave is real and how much is hype in a new costume?
Both are true at once, which is why the question needs method rather than opinion. What I look at is where money is actually being deployed, what has moved into production and stayed there, what regulators are formalizing, and where capability is compounding instead of being demonstrated once. My foresight work is grounded in operations research practice and stress tested adversarially, so a signal has to survive a deliberate attempt to break it before it reaches a recommendation. That review process is published as Crucible.
Are we actually behind our competitors on AI, or is it just talk at conferences?
Usually everyone is further behind than they sound and further ahead than you fear. What matters is not who announced a program, it is who changed a cost line, a cycle time, or a win rate. I benchmark on that basis, sector by sector, and I say plainly when the evidence is thin rather than filling the gap with narrative. A market radar engagement gives you a picture you can take to a board without borrowing anyone else's marketing to do it.
How do I know which signals to take seriously before I commit capital?
Test them for three properties. Is the signal observable in more than one independent source, is it accelerating or simply being repeated, and would you change a decision if it were true. Most conference noise fails the third test. My foresight practice runs signals through structured adversarial review before they become recommendations, which is deliberately slower and considerably harder to fool. Pricing is scoped to the engagement and to the coverage you need.
Where can I read something credible on the sovereign AI landscape in this region?
I publish the Sovereign AI Landscape for the GCC and Lebanon 2026 at michaeljoseph.ai/sovereign-ai-landscape-2026/, alongside a live monthly Sovereign AI Tracker that follows how the picture moves. Both are public, and both name sources so you can check the reasoning rather than accept a conclusion on trust. If you need the same discipline applied to your own sector, jurisdiction, or portfolio, that is a foresight engagement and it starts with a working session on the decision you face.
Discuss the scope
Discuss an intelligence need.
Share your objective, the decision you face, and your timeline. We can discuss the relevant work, the deliverables, and whether the engagement fits.
Contact MichaelRelated: Strategic foresight and AI · Market radar · Sovereign & institution-scale AI · All services