AI Washing vs Greenwashing: The 2026 Compliance Convergence

What Is AI Washing?

AI washing is the practice of overstating a product's, service's, or company's use of artificial intelligence to attract investors or consumers. The SEC's Emerging Technologies Task Force — established in 2025 and now part of the Cybersecurity & Emerging Technologies Unit (CETU) — uses the working definition: "material misstatements or omissions regarding a registrant's artificial intelligence capabilities, use, or impact."

The structural parallel with greenwashing is precise: both involve claims in marketing or investor communications that cannot be substantiated by the underlying operational reality, and both gain enforcement traction because the claims are material to consumer or investor decisions.

The Greenwashing Parallel

The Fortune / Baker McKenzie analysis (Dyan Finguerra-DuCharme, 23 April 2026) lists the structural similarities:

Both ESG investing and AI investing have grown into large, liquid capital pools where disclosure determines valuation multiples. That is the condition that makes mischaracterization actionable under securities law — and, in the EU, under the Unfair Commercial Practices Directive 2005/29.

SEC Enforcement Actions (2024–2026)

Source: Baker McKenzie Global Litigation News — SEC AI washing and Fortune — Inflated AI claims under fire (23 April 2026).

EU and UK Exposure Points

In the EU, AI-washing claims made to consumers fall squarely inside the Unfair Commercial Practices Directive 2005/29 (as amended by the ECGT Directive 2024/825). Two prohibitions apply:

• Article 6 UCPD — Misleading actions. Any false representation about the main characteristics of a product ("AI-driven") or about the results to be expected from use ("99% accuracy") is a misleading action.
• Article 7 UCPD — Misleading omissions. Failure to disclose that a service is human-operated when presented as AI-driven is a misleading omission when that information is material to the average consumer's decision.

In the UK, the Advertising Standards Authority (ASA) has ruled on several AI-capability claims in advertising since 2024, and from 6 April 2025 the CMA has direct enforcement powers under the Digital Markets, Competition and Consumers Act 2024 — the same 10% of global turnover ceiling that applies to greenwashing.

The EU AI Act (Regulation (EU) 2024/1689) adds a further layer: from 2 August 2026, providers of general-purpose AI models must publish a summary of training data and comply with transparency requirements. Unsubstantiated capability claims by high-risk AI deployers can trigger fines up to €35 million or 7% of global turnover under Article 99.

Where AI Claims and Green Claims Collide

A growing subset of corporate communications makes both sustainability and AI claims in the same breath:

• "Our AI-optimized supply chain reduces carbon emissions by 30%."
• "Climate-smart data centers powered by proprietary machine-learning cooling."
• "AI-driven ESG scoring delivers verified sustainability metrics."

Each of these statements has two substantiation tests: the AI capability must be real, and the environmental outcome must be measured and disclosed. The Guardian's February 2026 report "Claims that AI can help fix climate dismissed as 'greenwashing'" catalogued dozens of similar claims by tech companies, finding that few could be substantiated on either axis. A Fieldfisher alert (20 June 2025) identified this overlap as "the fastest-growing category of green claims risk."

For regulators, dual-claim statements are easy targets: a single investigation can deliver two violations from the same disclosure. The SEC has already begun coordinating AI-washing and climate-risk disclosure reviews through its CETU task force.

Practical Compliance Mapping

1. Combine your green-claim review and AI-claim review. A single review workflow should check every consumer-facing claim on a marketing page for (a) environmental substantiation and (b) technology-capability substantiation. Treat both as the same class of risk.
2. Apply the ECGT substantiation test to AI claims. If a claim's evidence is not accessible to the average consumer (no link, no published methodology, no third-party verification), the claim fails the same test the ECGT applies to "eco-friendly" terms. Rewrite or remove.
3. Disclose human-in-the-loop architectures. If humans make the decisions an AI is marketed as making, that must be disclosed — per UCPD Article 7 and SEC disclosure rules. Our free scanner can be extended to AI-claim audits by adding "AI-powered," "machine learning," and "autonomous decision" to the term list.
4. Price exposure at the intersection. The overlap zone is the highest-risk category: EU 4% widespread-infringement floor, UK 10% DMCC cap, and SEC unlimited civil penalties all apply simultaneously.