AI copywriting tools are now standard in affiliate email programs. Generate ten subject lines, five body variants, and a handful of "results" statements in the time it used to take to write one email. The speed is real. So is a problem most affiliate programs haven't caught up to: AI models don't know what's true, they know what sounds persuasive, and those aren't the same thing.
If an affiliate's AI tool writes "clients save an average of 40%" or "guaranteed approval" and nobody can produce data behind that number, you have a Section 5 problem, not an AI problem. The FTC doesn't grade claims on a curve because a machine wrote them.
The Standard Doesn't Change for AI
Section 5 of the FTC Act prohibits unfair or deceptive acts or practices in advertising. That includes a longstanding requirement: claims about a product or service need competent and reliable evidence behind them before they go out, not after someone asks.
The FTC has been explicit that AI doesn't get a pass here. Since launching Operation AI Comply in September 2024, the agency has brought more than a dozen enforcement actions over inflated or unsubstantiated AI-related claims, a pattern that continued through 2025 and into 2026. In one recent example, the FTC settled with Growth Cave for $48.6 million over allegations that included unsubstantiated automation claims paired with fake testimonials and manipulated reviews. Separately, the agency settled with Workado in 2025 over unsubstantiated performance claims for an AI detection tool, and with Cleo AI for $17 million over misleading claims about cash-advance amounts and timing.
None of these cases were about affiliate email specifically. That's the point worth sitting with: the FTC's claim-substantiation standard applies everywhere claims appear, on a website, in an ad, or in a promotional email an affiliate sent on your brand's behalf. There's no email carve-out.
Where This Actually Bites in Affiliate Programs
The exposure isn't really about brands writing false claims on purpose. It's about volume outrunning review.
When one affiliate submits one email for approval, someone reads it, checks the numbers, and signs off. When that same affiliate uses AI to generate ten variants and rotates them mid-campaign, sub-affiliates included, the review process built for one email a month doesn't scale to ten emails a week. AI tools are good at producing specific-sounding numbers ("increases conversion by 27%") that feel substantiated because they're precise, even when nothing backs them.
A useful gut-check before any AI-assisted claim goes out: would this statement survive if you had to hand a regulator the data behind it right now? If the honest answer is "we'd have to go find out," that claim isn't ready to send.
What This Means for Your Program
You don't need to ban AI copywriting tools. You need a substantiation checkpoint that doesn't disappear when content volume goes up. A few things worth building into your process:
- Flag numeric and comparative claims for a substantiation check, separate from the general creative-approval pass. "Saves you money" is different from "saves you 40%," and only one of those needs a data source behind it.
- Treat AI-generated variants the same as human-written ones. If your approval workflow only checks the first version an affiliate submits, AI-generated variants 2 through 10 are shipping unreviewed.
- Ask affiliates where a specific number came from, not just whether it's accurate. "Accurate" and "substantiated" aren't the same thing under Section 5, you need to be able to point to the evidence, not just believe the claim.
Pre-send review catches what affiliates submit for approval. It doesn't catch what changes after approval, or what a sub-affiliate adds downstream. That's where post-send visibility matters, and it's a gap most compliance workflows built for a slower content cycle weren't designed to close.
LashBack's BrandAlert lets you search a database of hundreds of millions of real-world commercial emails to see what's actually circulating under your brand's name right now, including claims you never reviewed. ComplianceMonitor picks up where that leaves off, watching affiliate-sent email in real time so unsubstantiated claims get caught before they compound into a Section 5 problem. Request a demo to see what your partner network is actually sending.
