The Incorrect Assumption Of 'Not American' In AI Standards

📊 Full opportunity report: The Incorrect Assumption Of 'Not American' In AI Standards on ThorstenMeyerAI.com — validation score, market gap, and execution plan.

TL;DR

European perceptions of AI sovereignty have shifted, mistakenly equating ‘not American’ companies with being outside US legal jurisdiction. Canada’s legal protections and international agreements complicate this view. The distinction impacts procurement and regulation strategies.

European policymakers have increasingly relied on the assumption that AI companies not incorporated in the US are outside US legal reach, particularly regarding the CLOUD Act. This assumption influences procurement and sovereignty debates, but it is based on a misunderstanding of legal and jurisdictional realities, especially concerning Canadian companies like Cohere.

It is confirmed that the CLOUD Act applies only to US-incorporated providers and their subsidiaries, meaning a Canadian-incorporated company like Cohere is not directly subject to US data access laws. Canada has not signed a bilateral CLOUD Act agreement with the US, and Canadian courts have explicitly rejected the third-party doctrine, which would allow US authorities to access Canadian data via US standards.

Furthermore, Canada’s foreign intelligence laws and oversight mechanisms are designed to protect Canadians’ data from foreign surveillance, including restrictions on targeting individuals in Canada. Canada’s adequacy decision from the European Commission, granted in 2002 and reaffirmed in 2024, makes data transfers from the EU to Canada legally valid, but only within certain limits and under specific legal frameworks.

However, critics argue that this legal distinction is often misused as a proxy for sovereignty, leading to an oversimplified view that ‘not American’ equals ‘outside US jurisdiction,’ which is not accurate in legal or practical terms.

At a glance
analysisWhen: developing; recent European policy disc…
The developmentEuropean policymakers have implicitly shifted from ‘incorporated in the EU’ to ‘not American’ as a proxy for sovereignty, based on flawed assumptions about legal reach and nationality.
The Wrong Test — Reality Check
AI Dispatch · Reality Check · 16 July 2026

The wrong test: “not American” is not a sovereignty standard

In one press conference, European sovereignty changed definition — from “incorporated in the EU” to “not incorporated in the US” — and nobody asked whether the second is a test or merely a proxy. It’s a proxy. Proxies fail at the edges. The edges are where procurement lives.

✓ First, what’s true — the Canadian case is stronger than critics allow

The CLOUD Act genuinely doesn’t reach Canadian incorporation. Canada has no CLOUD Act executive agreement — negotiating since March 2022, nothing finalized. And the Supreme Court of Canada (R. v. Spencer, R. v. Bykovets) explicitly rejected the US third-party doctrine. On several dimensions Canada is more protective than the US. This is not a hit piece.

The Five Eyes fact, stated precisely

UKUSA (1946): NSA · GCHQ · CSE · ASD · GCSB. CSE’s oversight is real — ministerial authorization, an independent Intelligence Commissioner (a retired judge) who can block, NSIRA review. Now read the operative restriction:

“CSE is prohibited by law from targeting the private information of Canadians, or any person in Canada.”

The protection is national and territorial. Europeans are neither.

Not an accusation — architecture. It’s structurally why Safe Harbor fell: protections protect the home nationals.

The adequacy gap nobody mentions

Canada has adequacy since 2001/2002 (Decision 2002/2/EC). But its scope is PIPEDA-only — employee data largely excluded; Alberta/BC/Quebec regimes never got adequacy; Quebec’s was withdrawn in 2014.

It was assessed against PIPEDA’s commercial framework — not against Canada’s intelligence laws or Five Eyes participation.

That’s the same hole the CJEU punched through Safe Harbor. In fairness: the Commission did examine public-authority access and found redress “accessible to non-Canadian nationals.” That clause is the best argument Canada has — and NSIRA is largely classified. Unsettled, not resolved.

⚠ The nexus problem — incorporation is not the test

US courts have been clear for 40 years: Bank of Nova Scotia — American courts enforce subpoenas against entities subject to US jurisdiction even where compliance violates foreign law, and fine for refusal. Jurisdiction attaches to presence and activity, not the incorporation certificate. So corporate pledges to “resist” are sincere and legally insufficient. And Canadian exposure creeps through ordinary commercial expansion:

BCE bought Ziply Fiber (US) Aug ’25 TELUS — 1,600+ US staff Shopify — 57% of txns in US; NY principal executive office None changed nationality. All changed nexus. So: what US nexus does Cohere have? Customers · ops · Microsoft partnership · US investors · a likely US listing. Nobody has asked.
The honest hierarchy — three standards, ranked by what they actually protect
✕ A proxy
“Not American”
Fails on nexus, fails on Five Eyes statutory architecture, fails when the ally’s interests diverge — and fails silently, because nobody’s measuring. This is what Europe just adopted.
◐ A test
“EU-incorporated”
SecNumCloud’s 24%/39% cap — narrow, arithmetic, checkable from a shareholder register. Also undeniably protectionist. Both true. What Europe already had — and just stepped back from.
✓ An architecture
Open weights · your keys · air-gappable
Requires trusting no jurisdiction, no ally, no election result, no executive directive. The only posture that survives every question below.
Europe just moved from the second to the first — and called it progress.
✓ The right test — enforceable, auditable control
1Who can compel you, under what standard, with what judicial review?
2Is there redress for a non-national? (US–UK/AU deals create none)
3What’s your nexus — not your incorporation?
4Who holds the keys, and can they be compelled to produce them?
5Can you leave, and how fast? (12–18 months of exit work)
6Can it be air-gapped?
Notice what happens down the list: the questions stop being about jurisdiction and start being about architecture. That’s not an accident — that’s the finding.
The take

The Five Eyes question isn’t “is Canada spying for America” — that’s the tabloid version, it’s unsupported, and it’s a distraction. The real question is duller and more damaging: why is Europe using nationality as a substitute for measurement? Because a proxy is cheap and a test is expensive. “Not American” lets you approve the deal, satisfy the minister, and skip the register, the nexus, the redress. It produces a press release. It does not produce protection. Every sovereignty claim here is a jurisdictional bet — that a legal system, an alliance and a political mood hold for the life of your data. The Canadian bet is genuinely better than the American one. It’s still a bet. The only positions that don’t require one are where you hold the weights and can pull the plug. If the answer is “well, they’re not American” — you haven’t been given a standard. You’ve been given a mood.

Sources: CSE’s own published material (UKUSA, mandate, Intelligence Commissioner, NSIRA, the targeting prohibition); IAPP, CIGI, Dentons, McMillan (Canada’s adequacy scope, PIPEDA limits, Quebec 2014); Barry Appleton, “Whose Law Governs Canadian Data?” (Balsillie Papers/SSRN 2026) & Citizen Lab Feb 2025 (Spencer/Bykovets, stalled CLOUD Act talks, Bank of Nova Scotia, UK’s 20,000+ requests, remedial no-man’s land, BCE/TELUS/Shopify nexus, US NSS & AI Action Plan). Some Five Eyes/GDPR analysis in circulation originates with vendors selling EU-hosted alternatives — read accordingly. Procurement & policy analysis, not an allegation of misconduct. Not legal advice.
thorstenmeyerai.com

Implications of Misjudging ‘Nationality’ in AI Legal Jurisdiction

This misconception influences European procurement policies and sovereignty assertions, potentially leading to flawed decisions based on an inaccurate understanding of jurisdictional risks. It risks oversimplifying complex legal realities and could hinder effective regulation and international cooperation in AI development and deployment.

Understanding that Canadian and other non-US companies are not automatically outside US legal reach is crucial for policymakers, companies, and regulators to make informed decisions and avoid strategic errors based on false assumptions about jurisdictional immunity.

Practical AI Governance: Building a Program for Oversight and Strategy

Practical AI Governance: Building a Program for Oversight and Strategy

As an affiliate, we earn on qualifying purchases.

As an affiliate, we earn on qualifying purchases.

Legal and Geopolitical Factors Shaping AI Jurisdiction Claims

The legal architecture around US data access laws, particularly the CLOUD Act, applies primarily to US-incorporated entities, not foreign companies with Canadian or European incorporation. Canada’s legal protections, oversight mechanisms, and international agreements, including the EU’s adequacy decision, create a complex landscape that resists simple categorization.

Recent European discussions have conflated ‘not American’ with ‘outside US jurisdiction,’ partly driven by political and sovereignty concerns. However, legal experts and courts, including Canada’s Supreme Court, have established that jurisdiction and legal protections are more nuanced, especially regarding foreign intelligence laws and international data transfer agreements.

This evolving narrative reflects broader geopolitical shifts and the importance of precise legal understanding in AI regulation and procurement.

Legal and Political Uncertainties in Jurisdictional Assumptions

It remains unclear how European policymakers will adjust their understanding of jurisdictional risks in AI procurement and regulation, especially as legal cases and international agreements evolve. The extent to which ‘not American’ will continue to serve as a proxy is uncertain, and ongoing negotiations and legal rulings could reshape the landscape.

Future Developments in Jurisdictional Clarity and Policy

European policymakers are likely to refine their understanding of jurisdictional risks, moving away from simplified proxies like ‘not American.’ Legal clarifications, international negotiations, and court rulings in Canada, the US, and Europe will shape future policies. Companies and regulators should monitor developments in US-Canada data-sharing agreements, court decisions on foreign surveillance, and EU policy adjustments.

Key Questions

Does US law automatically apply to Canadian or European AI companies?

No. US laws like the CLOUD Act apply primarily to US-incorporated entities and their subsidiaries. Canadian and European companies are generally outside US jurisdiction unless they have US-based operations or subsidiaries.

Can European companies avoid US jurisdiction by not being American?

Not necessarily. Jurisdiction depends on legal incorporation, operations, and international agreements, not just nationality. Relying solely on ‘not American’ as a proxy is misleading.

How does Canada protect data from US surveillance?

Canada’s legal framework explicitly restricts foreign intelligence agencies from targeting Canadians and includes oversight mechanisms. Its EU adequacy decision also limits data transfer scope.

Why do European policymakers believe ‘not American’ is sufficient for sovereignty?

They often see jurisdictional sovereignty as tied to legal nationality, but this oversimplifies complex legal realities and international data-sharing frameworks.

What should companies consider when evaluating jurisdictional risks?

They should analyze legal incorporation, international agreements, oversight mechanisms, and court rulings—beyond just the company’s nationality.

Source: ThorstenMeyerAI.com

This content is for general information only and is not financial, tax or legal advice. Consult a qualified professional for decisions about your money.
You May Also Like

South Korea’s ‘Ant’ Army Is Driving an AI Market Frenzy

South Korea’s retail investors, known as the ‘Ant’ Army, are fueling a rapid rise in AI-related stocks, creating a market frenzy that analysts are watching closely.

VigilSAR Benchmark: There Is No Best Model

VigilSAR Benchmark reveals there is no universally best AI model; rankings vary based on deployment needs, emphasizing trustworthiness and compliance.

Jack Clark Says It Out Loud — Reading the Co-Founder’s 60%/2028 Estimate on Automated AI R&D

Anthropic’s co-founder Jack Clark publicly estimates a 60% probability that autonomous AI system development could occur by 2028, signaling significant policy implications.

Two Channels: How the Pentagon Just Split Frontier-AI Procurement in Half

The Pentagon splits its AI procurement into two separate channels, placing Anthropic in a strategic, non-redundant segment, not excluded from federal contracts.