AI Marketing Claims Policy: Truthful Advertising and Prohibited Representations

The Problem of AI Marketing Exaggeration

AI marketing frequently overstates capabilities. Claims like "AI-powered accuracy," "intelligent automation," and "self-learning system" create expectations that products may not meet. When these claims influence purchasing decisions, they trigger consumer protection law. When they describe AI systems deployed in high-risk contexts, they may also create EU AI Act compliance issues.

The EU Unfair Commercial Practices Directive (2005/29/EC) prohibits misleading commercial practices, including false or deceptive claims about product characteristics, results to be expected, and the nature of the product. AI-related claims fall squarely within these prohibitions.

Regulatory Framework for AI Advertising

Prohibited AI Marketing Claims

The following claim categories should be prohibited in marketing materials:

• Unsubstantiated accuracy claims ("99% accurate") without specifying the test conditions, dataset, and metric used
• Autonomy overstatement ("fully autonomous," "no human needed") for systems requiring human oversight
• Capability exaggeration ("understands context," "thinks like a human") using anthropomorphic language that misrepresents technical function
• Privacy claims ("your data is never used for training") that contradict actual data processing practices
• Regulatory compliance claims ("fully compliant with the EU AI Act") before formal conformity assessment
• Performance guarantees that the system cannot reliably deliver in production conditions

Substantiation Requirements

Every factual claim about AI capability must be supported by documented evidence. For accuracy claims, provide the test dataset description, evaluation metric, and confidence intervals. For performance claims, provide benchmark conditions that reflect real-world deployment. For comparison claims, use equivalent testing conditions across compared products.

Maintain a claims substantiation file that links each marketing statement to supporting evidence. Update this file whenever marketing materials change or when system performance data is updated.

Review and Approval Process

All AI-related marketing materials should undergo review by: the product team (to verify technical accuracy), the legal team (to assess regulatory compliance), and the compliance team (to check alignment with internal policies). Establish a pre-publication checklist covering: claim substantiation, prohibited terms, transparency disclosures, and consistency with product documentation.

Competitor Claims and Comparative Advertising

The Comparative Advertising Directive (2006/114/EC) permits comparative advertising under specific conditions: comparisons must be objective, verifiable, and not misleading. When comparing AI systems, use standardized benchmarks and disclose testing methodology. Avoid claims about competitor AI systems based on assumptions rather than verified testing.

Remediation for Non-Compliant Claims

When non-compliant AI marketing claims are identified, implement immediate correction. Remove or modify the claim in all channels. Notify customers who may have relied on the claim in purchasing decisions. Document the corrective action. Conduct root cause analysis to prevent recurrence. National advertising standards authorities (ASA in the UK, ARPP in France, Werberat in Germany) may require public corrections.

Training for Marketing Teams

Train marketing personnel on AI literacy sufficient to evaluate claims critically. Provide clear guidelines distinguishing between permissible and prohibited language. Create a glossary of approved and prohibited terms. Conduct annual refresher training and update guidance as AI capabilities and regulations evolve.

This article is for informational purposes only and does not constitute legal advice. Regulatory requirements change frequently — verify current rules with official sources. Built by Sawai Gyoseishoshi Office, Hiroshima, Japan.