Strategic Research Series
Operational AI, sovereign technology, and the future of delivery, read for the people who run institutions in the Kingdom and the Gulf. Written in the field, not from a distance.
Saudi banks under-earn on fee income by roughly twenty points. Four levers, three scenarios, one number for the board.
Sovereign AIThe Kingdom is buying capacity faster than almost anyone. The scarce asset is governed, auditable delivery.
CompliancePDPL, NCA, SDAIA, and CST are one operating surface, designed in from the first decision.
Bank ScoreboardThe same gap that hides in Saudi banking reappears in insurance and on the exchange.
Vision 2030Announcements were the easy part. 2026 is when the Kingdom finds out who can operate what it has bought.
Gulf Macro84% bought AI. 11% earn from it. The Gulf that closes the gap between buying tools and building capability wins the decade.
Islamic FinanceThe same rules engine pointed at the same balance sheet. AI just made it cheap to run, for the youngest and largest pool of values-constrained capital on earth.
Sovereign AIThe Gulf is not capital buying its way into AI. It is cheap energy that AI is buying its way into. The cheque is visible; the cost curve is what lasts.
Gulf MacroOne policy fills the entry-level rung with nationals. AI is deleting that rung. The quota has to start counting judgment, not chairs.
The deliverable era is ending. Advisory reprices from fee-for-service to fee-plus-upside.
PricingWhen effort no longer equals value, the commercial model has to change with it.
Delivery ModelThe case for delivering from inside the operation, in Arabic, on Saudi ground.
Services FinanceMachine-readable scope turns a consulting pipeline into a balance-sheet asset, and eventually a tradeable instrument.
Services FinanceObservability is not assignability. A follow-up on what the structured-finance desk got right, and where the real prize sits.
Private EquityThe engagement was never the asset. The asset is the office that decides whose engagement is worth backing.
ConsultingThey automated away the very work that signalled they were worth the fee. Read the layoffs as a confession, not a strategy.
The gap between what AI agents can do and what an organisation can verify they did.
MarketplacesWhen the buyer is an agent, the winner has the cleanest data and the most reliable settlement, not the best search.
InfrastructureHundreds of billions are buying compute. Who governs what the capacity produces.
Operating ModelAn AI that lives inside your company and remembers everything rewrites the economics of every firm. The one line you must not cross.
Sovereign AIStop grading Arabic-first models on their Arabic. Grade them on who sets their refusals, their frames, and their defaults.
GovernanceThe Gulf is not behind on institutions. It is unencumbered by them. The lasting leapfrog is in the design of governance, not the model.
AI solved the information problem and worsened two others: the inability to commit and the inability to kill.
StrategyThe firms that dominate the next decade will not think faster. They will forget slower.
Operating ModelThe lever moves from where work happens to when. Wake-up advantage, if you can trust it.
StrategyProduction of belief has been commoditized. Distribution of belief has not.
FrontiersWhere durable advantage is being built before it becomes consensus.
ConvictionWhen everyone runs the same models and the information edge is competed to zero, the last thing left to price is the willingness to commit.
ConvictionConviction is credible only to the degree it cannot be reversed. A book of irreversible conviction is a tomb. The fix is a date stamp.
InvestingThe smartest trade in the AI cycle is to buy a boring company, not build a clever one. The durable edge is on sale.
StrategySuccession tests the family; AI dissolves the trade it was tested on. The two shocks are one filter, and only the institutionalized house survives both.
ConvictionA nation commits at a scale no company can match, which is exactly what makes it hard to undo. Conviction without an expiry calcifies. The fix is a designed half-life.
AI substitutes for cognitive labour itself. The skill-climb bargain breaks.
MacroThree proofs, a century apart, for why AI raises the price of the one thing it cannot do.
Market StructureCapital settles where the referee cannot join the game. Neutrality is the scarce asset a rival cannot buy, because it constrains the buyer.
Sovereign FundsThe sovereign fund stopped allocating and started operating. When capital and analysis go cheap, the only edge left is whether what you build stays standing.
Why the system you would design is never the system you get, and who collects the difference. A sixty-year-old law almost no strategist has read.
First PrinciplesWhy your most experienced people are the ones most likely to miss the regime change. An eye-tracking study of chess masters proved it.
First PrinciplesWhy well-run small firms get absorbed by worse-run large ones. It is physics, not strategy, and there is one move that beats it.
A city does not need another screen. It needs one operating picture and a way to act on it.
Digital HealthIn health, the trust layer is not optional. It is the product.
LogisticsYou can see the truck on a map. That is not the same as the decision it is about to miss.
EnergyThe plant floor is drowning in telemetry and starved of decisions.
MethodThe first conversation is a working brief, not a discovery call.
LanguageIn the Kingdom, Arabic-native operation is a moat, not an afterthought.