International Trade Commission Ai

The International Trade Commission conducted Section 337 investigations where complainants won findings of violation in 58% of cases in 2025 — up from 41% in 2024 — and AI is reshaping how both sides prepare for these high-stakes IP proceedings. Section 337 investigations move fast (typically 12-18 months to final determination), involve massive technical document sets, and require coordination between patent litigation, trade remedies, and customs enforcement. That's an environment where AI either saves your team or exposes it.

But here's the jurisdictional wrinkle that makes AI and the ITC uniquely complicated: Section 337 applies only to physical 'articles,' and the ITC generally lacks authority over intangible digital imports like electronically transmitted AI. That means AI-generated products, AI-as-a-service imports, and software delivered via the cloud may fall outside the ITC's reach entirely — creating a gap that sophisticated respondents are already exploiting.

Section 337 Investigations and AI Evidence

Section 337 investigations at the ITC are among the most document-intensive proceedings in U.S. IP law. A single investigation can involve millions of pages of technical documents, source code repositories, and prior art references. AI-powered document review tools provide a genuine competitive advantage — processing technical documents in hours that would take human reviewers weeks. But the ITC's evidentiary standards create unique requirements. Administrative Law Judges at the ITC evaluate evidence under their own procedural rules (19 C.F.R. Part 210), which differ from the Federal Rules of Civil Procedure. AI-generated evidence summaries and claim charts must be verified against these specific standards. A claim chart generated by AI that maps accused products to patent claims needs the same level of attorney verification as one prepared manually — the ITC ALJ will scrutinize it regardless of how it was created.

The Intangible Import Problem: AI Products and ITC Jurisdiction

Section 337 was designed for physical goods crossing borders, and the ITC's jurisdiction generally doesn't extend to intangible digital imports. This creates a growing gap as AI products increasingly arrive as cloud services, API calls, and electronically transmitted software rather than physical articles. Finnegan's analysis highlights this tension — complainants bringing Section 337 actions against AI-powered products must demonstrate that a physical article is being imported, which may require creative arguments about the hardware running the AI or the data storage devices involved. For respondents, the jurisdictional argument is a powerful defense: if your AI product is delivered entirely via cloud infrastructure without physical importation, the ITC may lack authority to issue an exclusion order. This jurisdictional question is unsettled and will likely require Congressional action or Federal Circuit guidance to resolve.

AI-Assisted Patent Analysis in ITC Proceedings

ITC proceedings compress litigation timelines dramatically compared to district court patent cases. Discovery, claim construction, and trial happen within 12-18 months, and extensions are rare. AI tools that accelerate patent analysis provide outsized value in this compressed timeline. Specifically, AI excels at prior art identification across multiple patent databases and languages — critical in ITC cases where respondents often involve non-U.S. companies and foreign-language prior art. AI-powered translation and semantic search tools can identify relevant prior art in Japanese, Chinese, Korean, and German patent databases that manual searches would miss or delay. Claim construction analysis is another high-value application. AI tools can parse prosecution histories, identify claim term disputes across related proceedings, and flag potential Markman hearing issues faster than manual review. In ITC proceedings where the claim construction hearing can determine the entire investigation's outcome, speed and thoroughness in this phase matter enormously.

Domestic Industry Requirement and AI Documentation

Every Section 337 complaint must establish a domestic industry — proving that the complainant exploits the patent through significant investment in plant and equipment, significant employment of labor or capital, or substantial investment in exploitation of the patent. In March 2025, the Federal Circuit ruled that the ITC's interpretation of the domestic industry requirement was overly narrow, expanding the range of activities that can satisfy this threshold. AI can assist in documenting domestic industry by analyzing financial records, employment data, and investment documentation to build the economic prong of the domestic industry case. For the technical prong, AI tools can map specific product features to patent claims, demonstrating that the complainant's products actually practice the asserted patents. This mapping exercise — tedious when done manually across a product portfolio — is well-suited to AI-assisted analysis.

Strategic Considerations for Managing Partners

ITC practice is concentrated among a relatively small number of firms — Finnegan, Fish & Richardson, Morrison & Foerster, and a handful of others handle a disproportionate share of Section 337 investigations. For managing partners at these firms, AI adoption isn't optional — it's a competitive requirement driven by compressed timelines and massive document volumes. The Trump Administration's expressed support for Section 337 and broader IP enforcement suggests investigation volumes may increase, potentially with new Commissioner appointments bringing different perspectives to institution decisions. Firms should invest in AI tools specific to ITC practice: patent analytics platforms with multi-language prior art search, e-discovery tools optimized for technical document sets, and claim chart generators that can process source code repositories. The firms that build these capabilities now will capture share as ITC practice grows. Those that rely on manual processes will struggle to compete on timelines and pricing.

The Bottom Line: The ITC's compressed timelines and massive technical document sets make it one of the highest-ROI environments for legal AI adoption. But the jurisdictional gap for intangible digital imports — including AI products delivered via cloud — creates novel defense strategies that firms must understand. With complainant win rates at 58% in 2025 and potential volume increases under the current administration, investing in ITC-specific AI capabilities is a competitive necessity, not a luxury.