The Briefing Table
our monthly briefing on the forces shaping markets, geopolitics, and corporate decision-making.
The Briefing Table – April 2026
Welcome to the April issue of The Briefing Table. This month’s edition covers four forces that defined the past few weeks and that will likely shape the strategic environment through the rest of 2026: the Iran war and its consequences beyond the oil price, the Supreme Court’s landmark tariff ruling, a U.S. economy absorbing both shocks simultaneously, and an AI governance gap that is closing in on its first major regulatory deadlines.
1) The Iran War: Beyond the Oil Price
The most significant development of the past week is the collision of simultaneous signals that point in opposite directions and remained unresolved after last night’s presidential address. Thousands of additional U.S. troops have arrived in the region, reportedly positioned for a potential seizure of Kharg Island or Iranian nuclear stockpiles. The administration has explicitly threatened to withdraw from NATO if European allies do not take a greater share of the burden in protecting Hormuz, which suggests a willingness to use the alliance itself as a bargaining chip. The economic consequences of the war were barely mentioned. Markets responded with oil prices up and futures down. Whether all this represents a strategy or the absence of one is genuinely unclear. What is clear is that resolution, when it comes, is likely to be abrupt, and, unlikely to restore the conditions that existed before the war began. (Readers looking for the full conflict scenario framework will find it in our recent special briefing; this section focuses on what has changed and on two exposure dimensions that are receiving insufficient attention.)
The first underappreciated exposure is competitive rather than operational. This shock is not landing symmetrically around the world. China holds crude reserves estimated at roughly 120 days of imports, its domestic pricing mechanisms insulate manufacturers from immediate pass-through inflation, and its twenty-year electrification bet is paying off precisely when its Western rivals are most exposed. Chinese manufacturers may actually gain a cost advantage in energy-intensive sectors as this shock persists, echoing what happened after the Russia-Ukraine war in 2022. McKinsey’s March 2026 Global Survey confirms the asymmetry in executive sentiment: respondents in Asia-Pacific, Greater China, and India lean more positive than negative about near-term conditions, while those in Europe and North America are the most likely to report worsening conditions and the least likely to expect improvement. The deeper irony: with China controlling more than 80% of global solar, wind turbine, and battery production capacity, the supply chain for the energy transition this conflict is accelerating runs predominantly through Beijing. For executives in energy-intensive industries, this is a present-tense competitiveness question, not a future-tense strategic concern.
The second is the semiconductor and AI infrastructure exposure. Qatar’s Ras Laffan shutdown has closed off a major source of helium, essential for chip fabrication, fiber optics, and MRI machines. The Gulf region supplies roughly half of global sulfur exports, the feedstock for the high-purity sulfuric acid required to manufacture clean silicon wafers. Iranian drone strikes have already damaged two AWS data centers in the UAE and a facility in Bahrain. Gulf sovereign wealth funds, which have been among the largest investors in AI infrastructure globally, are redirecting capital toward domestic reconstruction. The conflict is simultaneously disrupting the physical inputs, the cloud infrastructure, and the investment capital that underpin most companies’ AI buildout assumptions. Few strategy teams have modeled all three.
Bottom Line: Most energy cost exposure has now been identified. The less-modeled risks are China gaining ground in energy-intensive industries while Western rivals absorb the shock and supply chain disruption running through semiconductor inputs and AI infrastructure. Both warrant immediate audit. On scenario planning: do not treat a sudden ceasefire as a return to normal. Iran has demonstrated it can close the Strait; that capability does not disappear with a peace agreement. Real shipping and oil market disruptions would persist for weeks after any ceasefire, Iranian leverage over the waterway is now an established fact rather than a theoretical risk, and the regional power balance has shifted in ways that will outlast the conflict. The status quo ante is gone. Plan for a new baseline, not a restoration.
2) The IEEPA Ruling: The Tariff Wars Enter a New Phase
On February 20, the Supreme Court ruled 6 to 3 that IEEPA does not authorize the President to impose tariffs. Chief Justice Roberts held that tariff authority is “very clearly a branch of the taxing power” reserved for Congress. The decision invalidated the reciprocal tariffs and most of the tariff architecture built since Liberation Day. As we previewed in December, the ruling has created nearly as much new uncertainty as it resolved.
The administration moved within hours. Section 122 of the Trade Act of 1974 now imposes a 10% global tariff pushed toward the 15% ceiling, expiring July 24 unless Congress extends it. New Section 301 investigations were launched targeting 16 economies, including China, the EU, Japan, Mexico, and Vietnam, for structural excess capacity. An additional 60 investigations were launched for failures to enforce forced-labor bans. Section 232 tariffs on steel, aluminum, and autos remain fully intact. As a result, the effective tariff rate, at approximately 10.5%, remains at its highest level since 1943. The paradox of the ruling is that it has replaced a fast, legally fragile instrument with slower ones that are harder to challenge and longer-lasting in their effects. IEEPA tariffs could be imposed overnight and litigated immediately. Section 301 findings take 12 to 18 months to reach conclusions which are far more durable once they do.
Two pressure points deserve attention. Penn Wharton projects up to $175 billion in potential IEEPA refunds, but the Court left the mechanism to the Court of International Trade, which is managing nearly 2,000 pending cases; CBP cannot begin processing until mid-April at the earliest. Large importers are sitting on a potential windfall wrapped in procedural uncertainty, arriving in the middle of an energy crisis. Separately, the July 1 USMCA review deadline, overshadowed by the IEEPA drama, may prove more consequential for North American supply chains than the ruling itself, and the negotiations may prove more consequential for North American supply chains than the ruling itself. Canada has recently signaled a willingness to explore a North American “fortress” approach to energy and raw materials, but that posture has not softened Washington’s negotiation position. The realistic range of outcomes includes not just a managed renewal with modified terms but also collapse of the trilateral framework into separate bilateral deals, or large unilateral actions against Canada specifically, which remains more exposed than Mexico given its energy and resource dependencies. One further signal worth registering: McKinsey’s March 2026 Global Survey finds that geopolitical instability has now overtaken trade policy as the most-cited threat to company growth for the first time since March 2025. The legal and trade architecture remains unsettled, but executive attention has already moved on.
Bottom Line: Three simultaneous actions: file or preserve IEEPA refund claims before procedural windows close; map your supply base against the 16 Section 301 investigation targets; and begin USMCA scenario planning before July 1 forces a reactive posture. The tariff environment has not stabilized. It has restructured around a less predictable legal architecture, and that restructuring is arriving into an economy simultaneously absorbing an oil shock and a monetary policy dilemma.
3) The U.S. Economy: Stagflation Lite Is No Longer a Forecast
In November, we identified two Gray Rhinos preparing to charge: tariff pass-through to consumers and the risk of an AI investment bubble. The Iran war has now added a third. What unites all three is a common structure: their costs were largely deferred in 2025 through inventory pre-stocking, energy price normalization, and AI investment that has not yet faced a demand test. They are now arriving simultaneously, however, and without the buffers that cushioned last year’s shocks.
The headline numbers still look passable: GDP growth projected around 2%, unemployment at 4.4%, markets volatile but intact. The structure beneath those numbers is less reassuring. The wealthiest 10% of households now generate nearly half of all consumer spending, with Deloitte estimating 20 to 25% of that spending has been driven by the wealth effect from rising asset prices. That support disappears quickly in a sharp equity correction. Bank of America Institute data shows that younger Gen Z and Millennials consumers, who had finally begun outspending older cohorts on the back of easing rents and strong wage growth, have disproportionately high gasoline spending relative to their discretionary budgets. A sustained fuel shock will reverse recent momentum in retail, restaurants, and consumer electronics precisely in the demographics that had been driving it.
McKinsey’s March 2026 Global Survey on Economic Conditions captures how quickly the picture changed. Executives surveyed before February 28 were as optimistic as they had been in December, the most upbeat quarter of 2025. From February 28 onward, the share saying global conditions had worsened nearly doubled, and forward expectations flipped from net positive to net negative within 72 hours. The share citing supply chain disruptions as a top risk doubled in the same window. This is not a gradual deterioration. It is the kind of sentiment shift that arrives faster than quarterly planning cycles can accommodate, which is why the contingency planning argument in this issue is not about managing disruption but about matching its speed. There is, however, a complicating signal: at the company level, expectations remain primarily optimistic. Just over half of private sector respondents still expect demand for their products and services to increase in the next six months, and about six in ten expect profits to grow. That gap between a sharply deteriorating macro-outlook and continued company-level optimism, is a pattern worth watching. It tends to resolve in one direction only, and it is rarely the macro picture that turns out to be wrong.
The monetary policy picture has deteriorated materially. The Fed held rates steady on March 18 and penciled in at most one cut for 2026, down from the two cuts markets had been pricing before the war. Several analysts now think no cuts will be delivered this year. The ECB, facing a more severe energy exposure, has already postponed planned reductions and warned of stagflation risk in Germany and Italy. The Fed cannot address both sides of this dilemma at once: holding rates to contain inflation risks tipping a softening labor market into something worse; cutting to support growth risks entrenching energy-driven price pressures. Any capital allocation plan built on an assumption of two or more rate cuts in 2026 needs to be revisited. Chairman Powell’s term expires in May. As we analyzed last July, a successor perceived as politically accommodative could push long-term bond yields higher even as short-term rates fall, precisely the wrong environment for capital-intensive businesses.
Bottom Line: Goldman Sachs has raised its 12-month U.S. recession probability to 30%; Moody’s Analytics sits at 42%. Review capital expenditure plans and debt covenants. Build variable cost flexibility into the operating model. But the most important action may be the simplest: put your macro assumptions and your company forecast in the same room and ask whether they are consistent. McKinsey’s survey suggests most executives will find that they are not. That gap does not stay open. It closes and it is rarely the macro picture that moves.
4) Agentic AI: The Governance Gap Is Now a Board-Level Liability
In December, we cautioned against the tendency to go “all-in” on emerging technologies before the enabling infrastructure is in place, a pattern repeated across EVs, autonomous vehicles, fuel cells, and telematics. That warning applies with even greater urgency to agentic AI, with one important difference: the automotive industry’s overcommitments primarily affected capital budgets. AI governance failures will affect customers, regulators, and reputations simultaneously, and the regulatory deadlines are no longer abstract.
The shift happening in 2026 is not simply more AI. Generative AI assistants merely make recommendations. Agentic systems act on them, initiating transactions, sending communications, modifying data, and orchestrating workflows across enterprise systems without human review at each step. Only 21% of enterprise leaders report having a mature governance model for autonomous agents. Meanwhile, 65% of AI tools inside enterprises operate without IT oversight, adding an average of $670,000 to the cost of a data breach. The AI running inside your organization today is almost certainly not fully visible to the people responsible for managing its risks.
The governance gap persists not because executives are unaware of it but because the standard response - approving AI tools through procurement and IT review - is structurally mismatched to the actual risk. Approving a tool is not the same as governing what it does. With generative AI assistants, that distinction was manageable: the human read the output and decided what to do with it. With agentic systems, the human who authorized the agent is typically nowhere near the workflow when the agent acts. The accountability gap is not a process failure. It is an architectural one.
Three regulatory deadlines are now inside any responsible planning horizon. The EU AI Act reaches full general application August 2. Colorado’s AI Act takes effect June 30, with specific obligations around algorithmic discrimination. California’s generative AI transparency requirements are already active. The FTC’s “Operation AI Comply” in 2025 established that regulators expect documented controls and technical safeguards, not aspirational ethics statements. The governance question boards should be asking is not whether to govern AI but at what level: governance applied at the tool level, approving AI systems, misses the risk. The risk is in what agents do, not what they are, which means governance needs to be built into the workflows agents operate within, with clear accountability for each action they can take.
Bottom Line: The governance gap in enterprise AI today is where corporate cybersecurity governance was in 2003: widely acknowledged, but rarely addressed with appropriate seriousness, until high-profile failures made inaction untenable. The firms that built the infrastructure before the breaches gained lasting competitive and reputational advantages.
Key Take-Aways
• Model the full exposure, not just the oil price. Map every input that transits the Strait of Hormuz or originates in the Gulf, including indirect dependencies like helium, sulfur, and industrial gases. Most companies have done this for crude oil. Few have done it for the second- and third-order inputs that are quietly closing semiconductor and AI supply chains.
• The tariff regime has restructured, not stabilized. File IEEPA refund claims promptly, map your supplier base against the 16 Section 301 targets, and begin USMCA planning before the July 1 deadline. The effective tariff rate remains near an 83-year high, and the administration has demonstrated it will find new instruments whenever old ones are removed.
• Your macro assumptions and your company forecast are probably inconsistent. McKinsey’s March survey finds executives remain primarily optimistic about their own firms’ demand and profits even as their macro outlook has deteriorated sharply. That gap tends to close in one direction. Audit your company-level plan against the macro picture explicitly, not as a separate exercise but in the same room, with the same set of assumptions, at the same time.
• Govern AI at the workflow level, not the tool level. The risk is in what agents do, not what they are. Commission a governance review that covers agentic systems specifically, maps accountability to individual agent actions, and meets the EU AI Act, Colorado, and FTC standards. The August deadline is the forcing function.
• Build contingency plans in both directions. A rapid negotiated off-ramp is as plausible as further escalation, and it could arrive with virtually no warning, as last Monday’s 14% oil price swing demonstrated in both directions inside a single trading session. Scenario planning should be explicit about both endpoints and the speed at which transitions can occur. Companies that prepare only for continued disruption will be as wrong-footed by a sudden normalization as those that failed to prepare for the disruption itself.