New U.S. export controls and tariff rulings raise costs for AI projects as February jobs soften

The Commerce Department and U.S. Trade authorities on Feb. 26 issued a package of export controls and tariff rulings that raise near-term costs for corporate AI programs, while Labor Department data the same day showed cooling in the jobs market that could blunt immediate wage pressure and hiring. Together the moves create a sharply different calculus for firms deciding whether to accelerate AI hiring and capital spending.

The trade actions broadened export restrictions on high-end AI chips and development tools and increased provisional tariffs on certain semiconductor manufacturing equipment by 10 percentage points, a change the government estimates affects roughly $3.2 billion in annual imports. The administration also added 15 foreign entities to a Commerce export‑control list, tightening access to chipmaking tools and advanced processors that firms use in large language model training and inference clusters. Officials framed the measures as protecting national security of critical supply chains, but companies say they raise compliance costs and delay shipments.

On the labor front, the Labor Department reported that nonfarm payrolls rose by 150,000 in February and the unemployment rate ticked up to 4.1 percent, while average hourly earnings increased 0.3 percent month over month. Job openings, tracked in the Government’s monthly survey, continued to decline and stood at 8.9 million, down from a peak the prior year. The combination of slower hiring and modest wage growth reduces immediate inducements for firms to expand payrolls, particularly in expensive AI talent markets.

For firms building AI systems, the twin signals are practical and immediate. Higher tariffs and export constraints lift the cost and lead time for acquiring specialized chips and lithography equipment. That forces many companies to choose between pausing model training, shifting workloads to less-efficient cloud providers that still have compliant hardware, or paying up for scarce domestic stock. The resulting uncertainty is prompting some CFOs to delay planned head count increases for AI engineering and infrastructure teams until supply and tariff risk clears.

Financially, a 10 percentage point tariff on $3.2 billion of imports translates to roughly $320 million of additional direct duty annually, before pass-through, compliance and logistics costs. For large tech firms spending hundreds of millions on data-center hardware, that is meaningful. Smaller firms and AI startups, which rely on third-party hardware procurement, face higher marginal costs and tighter cash flow, making outside funding and hiring less attractive.

From a policy perspective, the moves reveal a trade-off for Washington between structural supply‑chain security and short-term competitiveness. Greater protection of critical technologies aims to reduce long‑run strategic dependencies, but in the near term it raises costs for U.S. businesses competing globally in AI. The softer labor data gives policymakers room to tighten trade without immediately fueling broader wage shocks, but it also leaves workers in disrupted sectors exposed if firms curb hiring.

Longer term, these developments accelerate trends already visible in 2024 and 2025: partial reshoring of chip supply chains, increased vertical integration by cloud providers, and a stronger premium on software optimization over brute-force compute. For corporate leaders weighing AI investments, the smart money will be on flexibility: contracts with diverse hardware suppliers, staged hiring tied to demonstrable compute availability, and capital plans that assume higher near-term costs but potential long-run benefits from domestic supply resilience.