Executive summaryTechnology tariffs affect more than the landed cost of finished electronics, which is why technology trade compliance now reaches the board. Tech imports and exports depend on global supply chains of raw materials, semiconductors, specialist manufacturing equipment, components, and final assembly. Understanding this trade architecture is essential to assessing the true cost of innovation. |
Key insights
|
Innovation has a physical footprint.
A technology business might think its value resides in IP, engineering code, software, AI or scientific knowledge. But commercialising that innovation depends on semiconductors fabricated overseas, specialist manufacturing equipment, critical raw materials, networking hardware, sensors, servers, components, and assembly operations spread across multiple jurisdictions.
So the real cost of technology imports and exports isn’t simply the tariff attached to the finished product, but the cumulative trade architecture beneath that product… and the commercial decisions that architecture necessitates.
Why this mattersTechnology supply chains can cross several borders before a finished product reaches its customer. When tariffs change, the consequences can travel backwards through sourcing and manufacturing – affecting landed cost, supply chain design, pricing and, ultimately, market competitiveness. |
Your product has a passport. Its supply chain has dozens.
Consider the physical infrastructure behind artificial intelligence.
The software may operate through cloud computing, but delivering it requires advanced processors, memory, servers, networking equipment, cooling systems, energy infrastructure and data centres. In turn, the machines that produce those chips depend on specialist engineering, materials and semiconductor equipment sourced through another global supply chain.
The principle applies across modern technology: smartphones, medical technology, robotics, quantum systems, communications infrastructure and advanced electronics are all developed across multiple countries.
Even a simplified technology supply chain might look like this:
| Layer | Trade dependency | Potential commercial exposure |
| Raw materials | Critical minerals and specialist materials | Tariffs, sourcing risk, availability |
| Components | Chips, memory, sensors and electronics | Classification, origin, import duties |
| Production | Fabrication and specialist machinery | Export controls, market access |
| Assembly | Cross-border manufacturing | Origin, tariffs, customs |
| Infrastructure | Servers, networking and data centres | Import cost, capacity, supply |
| Finished product | Hardware and connected technology | Tariffs, pricing, competitiveness |
“The tariff attached to the finished product is only the most visible layer of its trade exposure.”
What are technology tariffs – and who pays them?
A tariff is a duty imposed on imported goods: for a tech company, that might apply to a finished laptop, semiconductor, electronic component, manufacturing tool, or another physical input within the production process.
But the commercial impact is not as simple as:
10% tariff = customer pays 10% more.
The importer initially bears the customs liability. From there, businesses may absorb the cost, negotiate with suppliers, increase customer prices, change sourcing, alter manufacturing locations, or accept a reduction in margin.
US tariffs cost Apple approximately $800 million during the quarter ending June 2025. The firm then projected another $1.1 billion of tariff-related costs for the following quarter. At the same time, Apple was shifting production of US-bound products towards India and Vietnam to reduce its exposure.
“If tariffs can add $800 million to the costs of one of the world’s largest companies in a single quarter, trade exposure must be considered an active and commercial variable.”
Tariffs can redraw supply chains
In February 2025, HP announced that its financial outlook reflected increased US tariffs on China and the measures it was taking to mitigate them. Senior leadership expected that – by the end of its 2025 financial year – more than 90% of products sold in North America would be manufactured outside China.
“A lower tariff does not automatically mean a lower total cost. Supply diversification only creates value when the economics make sense after customs, logistics, and operational risk are included.”
That illustrates how far tariff exposure can reach. It has the potential to change:
- Where suppliers are located.
- Where products are manufactured.
- How much inventory businesses hold.
- And which production footprint remains commercially viable.
Relocation itself is not a frictionless exercise – new suppliers need qualification, manufacturing processes need replication, and logistics networks change. Quality and capacity have to be maintained.
Moving the factory doesn’t necessarily move the tariff
Country of origin is not simply whatever country appears on the final shipping label. For US customs purposes, determining whether manufacturing in a new location changes origin can depend on whether the processing amounts to a substantial transformation (broadly, whether an article emerges with a new name, character or use).
US Customs and Border Protection considers factors including component origin and the nature / extent of manufacturing on a case-by-case basis. CBP has previously ruled that certain Infineon semiconductor products assembled and packaged in China retained the origin of the country where the wafer’s front-end processing took place: Germany, Austria or Malaysia. The later Chinese manufacturing steps did not substantially transform the underlying transistor circuitry.
For leadership teams, the implication is this: moving the final assembly stage changes logistics. It does not necessarily change the customs outcome.
AI is digital – but its supply chain isn’t
The latest US approach to advanced computing chips illustrates the physical economics of innovation.
In January 2026, the Trump administration introduced a 25% tariff on certain advanced computing chips, including NVIDIA’s H200 and AMD’s MI325X. Yet the measure exempts specified uses including US data centres, R&D, startups and public-sector applications, alongside other uses judged to strengthen domestic technology capacity.
That distinction is revealing: the same semiconductor can receive different tariff treatment depending on how it enters the wider technology ecosystem.
Policy is therefore influencing not only the cost of importing a chip, but the incentives around where infrastructure, R&D, and manufacturing are built.
“The cloud still has to be built somewhere.”
AI may feel intangible, but its progress requires enormous volumes of physical infrastructure. Every server, processor, networking component and machine has an origin, classification, supply chain and route across borders.
Stress-test the trade architecture behind your innovation
Technology companies do not need to predict every future tariff. They do need to know where trade exposure sits before one changes. Leadership teams should be able to ask:
- Which components and raw materials create our greatest tariff exposure?
- Where does country of origin actually sit within the production process?
- Are our customs classifications robust?
- Which costs could we absorb – and which would fundamentally change product economics?
- What alternative suppliers or manufacturing routes exist?
- What happens to margin if tariff assumptions change?
The history of technology is told through inventions, software, and scientific breakthroughs. Commercial reality is less weightless.
The more sophisticated innovation becomes, the more important it is to understand the physical trade architecture beneath it. Knowing what technology costs to import or export means understanding the global system that made the product possible.
| Borders For the Boardroom:
Episode 3 | Being proactive at the border Why your trade strategy needs to be proactive, and how to build resilience in the new world of trade tariffs. |