The Short Answer: Yes, But the Old Rules Still Apply

Yes, you can use AI for marketing, and no jurisdiction covered by this research bans it outright. The catch is that AI-generated content has to clear the same bars as anything a human writes: truthful claims, cleared intellectual property, and the industry rules that already govern your ads.

The UK's advertising regulator puts it plainly. The ASA and CAP codes contain no AI-specific rules, but existing rules apply regardless of how content is generated, edited, or targeted. The same principle runs through US guidance: AI content must satisfy the truth-in-advertising standards, IP law, and state requirements that cover all marketing.

What changes with AI is speed and volume, not the nature of the risk. Cory Krasnoff, Senior Counsel at Uber, frames it as "just a technological tool," with IP questions that look a lot like the ones that came up when a designer pulled a photo off Google. That matters because enforcement is already active. The FTC's Operation AI Comply initiative has produced more than a dozen actions since 2024, including a $48.6 million settlement over inflated AI capability claims, and the agency has said marketing hype about AI will be measured against traditional substantiation requirements.

In practice, that means a few things stay non-negotiable no matter what tool you use:

• Every claim needs evidence behind it, including claims about what your AI can do.
• Someone accountable reviews the output before it ships. The ABA's position is blunt: AI-generated work must be reviewed and approved before release, and "grounding is not supervision."
• Disclosure rules depend on where you advertise. New York's synthetic performer law takes effect June 9, 2026, while the UK regulator warns that disclosure alone rarely fixes a misleading message.
• Your state obligations may not match your address. As Whiteford attorneys note, you don't need an office in Texas or California to be covered.

Most organizations still haven't built a formal review process for AI marketing content, which is where the real exposure sits. Worqd's own rule is simple: until real evidence is approved, we use clearly marked placeholders, and we don't invent revenue, logos, or testimonials. If you want to see how that plays out across paid ads, creative, and follow-up, book a growth call and we'll walk you through it.

Where the Real Risk Lives: Enforcement Is Active and Expensive

The question stops being academic the moment a regulator sends a letter. Enforcement is already active, and the price of getting AI claims wrong runs into eight figures.

The FTC's Operation AI Comply initiative has produced more than a dozen enforcement actions since 2024, targeting inflated or unsubstantiated AI capability claims. The headline case is the FTC's $48.6 million settlement with Growth Cave, where the agency challenged claims that AI software would automate "nearly 100%" of course-building work that users actually did by hand. The FTC's position is blunt: marketing hype about AI gets measured against traditional substantiation requirements.

That standard is exactly why your claims need evidence before your creative ships. It's also why Worqd's anti-fabrication policy exists: clearly marked placeholders until real evidence is approved, and a banned-words list that includes "guaranteed," "10x," and "unlimited." Those aren't stylistic preferences. They're risk controls.

The state picture matters just as much, because it's live right now:

• Texas TRAIGA and Illinois HB 3773 both took effect January 1, 2026.
• New York's synthetic performer disclosure law kicks in June 9, 2026, with escalating civil penalties for undisclosed AI-generated human likenesses in ads.
• California SB 53 applies its heaviest obligations to developers earning over $500 million annually.
• Connecticut's CART Act phases in from October 1, 2026 through January 2028.

Here's the part most businesses miss. You can be covered by a state's AI law based on where your customers are, not where your office is. As attorneys at Whiteford put it, you don't need an office in Texas or California to be covered, and a company serving a national market is often within scope. No comprehensive federal AI statute preempts any of this.

So when you're choosing a growth partner, ask what happens before content goes live. Most organizations still lack a formal legal review process for AI-generated marketing, which means the burden lands on you to verify that your provider has one.

What Compliant AI Marketing Actually Looks Like

Compliant AI marketing doesn't look like a disclaimer bolted onto the end of an ad. It looks like a system: people with authority reviewing work before it ships, risks sorted by severity, and every claim backed by evidence you could hand to a regulator.

Start with human oversight, because that's where most teams fall short. In one industry webinar poll, most respondents admitted they had no formal legal review process for AI-generated marketing content yet, or were still building one. The biggest challenge wasn't the technology. It was getting marketing to bring legal in early enough. Andra Dallas, Senior Counsel at PayPal, put it plainly: don't review things by email or Slack. "There has to be a system in place for review."

Uber's in-house team offers a working model worth copying. Rather than treating every piece of AI-assisted content the same, they maintain roughly 8-9 defined high-risk categories that trigger mandatory human review. Claims about regulated products, anything touching vulnerable populations, content that makes performance promises. Low-risk material moves faster; the risky stuff gets eyes on it before it goes live. For agencies working in verticals like legal, medical, or finance, this kind of triage isn't optional. It's the difference between a workflow and a liability.

Then there's substantiation. The FTC's "Operation AI Comply" has brought more than a dozen enforcement actions since 2024, including a $48.6 million settlement against a company that claimed its AI would automate nearly 100% of a task users still had to do mostly by hand. The FTC's position is that marketing hype about AI gets measured against traditional substantiation requirements. That's why a banned-words list matters. At Worqd, words like "guaranteed," "10x," and "game-changing" don't ship, and until real evidence is approved, creative uses clearly marked placeholders rather than invented results. It's a small internal rule with a very large external justification.

Disclosure is the trickier piece, because the US and UK are moving in different directions. New York has enacted the first law in the nation requiring businesses to disclose "synthetic performers" (AI-generated human likenesses) in commercial advertising, effective June 9, 2026, with escalating civil penalties. If you're running AI-generated UGC-style video ads, disclosure risk is now, as one advertising partner put it, "a live compliance concern."

The UK's regulator takes a different view. The ASA and CAP warn against treating disclosure as a checkbox, noting that "disclosure alone is very unlikely to mitigate the harm caused by a fundamentally misleading message." In other words: an AI label doesn't rescue an ad that misleads people. Marketers serving both markets have to navigate both philosophies at once.

If you're evaluating a growth partner, ask how they handle these four things. A good answer sounds like a documented review process, a risk-based triage model, evidence behind every claim, and a clear position on synthetic content disclosure. If the answer is vague, keep looking.

How to Vet an AI Marketing Provider's Compliance Practices

Most AI vendors look compliant on their website and fall apart the moment you read the actual paperwork. Legal teams who vet these tools regularly have learned this the hard way, and their process is worth copying before you sign with any AI marketing partner.

The core advice from practitioners is blunt: judge the privacy policy and data processing agreement, not the marketing page. One legal team's review process found that "most AI meeting tools fall apart the moment you read the actual privacy policy and DPA rather than the marketing page," and that "the shortlist that actually survives a data governance review is thinner than you'd think" (a practical comparison shared by legal teams).

Start with three questions, and get answers in writing:

• Does your data train their models? If the privacy policy and DPA don't clearly address training-data use and retention, the practitioner guidance is simple: keep looking.
• How long is data retained, and where does it go? Vague answers here are a red flag, especially if the tool touches your CRM or customer conversations.
• Is there a documented review workflow? Ask how a human checks AI output before it reaches your audience. The American Bar Association's guidance is that grounding in source material is not supervision, and every piece of AI-generated work needs human approval before release (the ABA's responsible AI checklist).

That last question matters more than most buyers realize. In one industry poll, most respondents said they have no formal legal review process for AI-generated marketing content yet, or are still building one (Axiom's compliance research). Andra Dallas, Senior Counsel at PayPal, put it plainly: "Don't just review things by email. Don't just review things by Slack. There has to be a system in place for review."

This gap is a real differentiator when you're choosing a partner. A provider who can show you a documented process, instead of a vague promise that "we review everything," is ahead of most of the market. It's also why we built Worqd's practices the way we did: a standing anti-fabrication policy that uses clearly marked placeholders until real evidence is approved, a banned-words list that outlaws "guaranteed" and "10x" claims, and explicit consent built into how we collect lead information. Those are the kinds of things you should be able to see, not just hear about.

One caution as you vet: ask for evidence of any certification a provider claims to hold. If a vendor can't produce it, treat the claim the same way the FTC treats unsubstantiated marketing hype, which it measures against traditional substantiation requirements. A partner who welcomes that scrutiny is usually the one worth keeping.

How Worqd Handles Compliance: Our Rules, In Plain Language

The research is clear that AI marketing is legal but held to the same truth-in-advertising rules as everything else, and that most organizations still have no formal review process in place, which is exactly the gap a provider should be able to describe in plain terms. Here is how Worqd answers that in its own published materials.

Worqd's anti-fabrication policy is the centerpiece. Until real evidence is approved, the team uses clearly marked placeholders, and nobody makes up revenue, conversion lifts, logos, testimonials, or named clients. That single rule maps directly onto the FTC's stated position that marketing hype about AI gets measured against traditional substantiation requirements, the same posture behind the $48.6 million Growth Cave settlement and the broader Operation AI Comply actions.

The banned-words list reinforces it. Words like 10x, guaranteed, revolutionary, game-changing, and unlimited are off the table, which is the internal version of the external standard.

Other practices round out the picture:

• B2B outreach is personalized and permission-aware, aimed at relevant accounts rather than template blasts.
• The booking funnel requires explicit consent to be contacted, and states that details are used only to prepare for the call.
• No sensitive form fields are sent to public analytics.

These choices line up with what legal teams actually vet vendors on. According to practitioner guidance on AI vendor reviews, if the privacy policy and data processing agreement do not clearly address training data use and retention, you keep looking. Most organizations still lack a formal AI review process, and the biggest complaint is getting legal involved early enough, per a webinar poll cited in compliance coverage. Since Worqd is a retainer-style growth partner rather than a platform, the rules it publishes are the compliance surface clients can actually inspect.

More demand. Faster follow-up. Better creative. If you want to see how those rules apply to your funnel, book a growth call.