Picture this: a marketing firm calls your small business promising a tool that "listens" to nearby smart devices to find customers already talking about products like yours, then targets them with hyper-relevant ads. It sounds like science fiction, except it's exactly what the Federal Trade Commission says three marketing companies sold to small business owners in 2026 — and the "AI-powered" listening tool didn't exist. The companies had simply bought email lists from data brokers and marked them up.
That case wasn't an outlier. It was the thirteenth enforcement action the FTC has filed since September 2024 under a sweeping initiative aimed at companies that slap "AI" on a product and let the label do the selling. If you've bought — or are shopping for — an "AI-powered" bookkeeping tool, sales assistant, hiring screener, or marketing platform, this crackdown is worth understanding before your next purchase decision.
What "AI-Washing" Actually Means
"AI-washing" describes marketing that exaggerates, misrepresents, or outright fabricates a product's artificial intelligence capabilities to make it sound more advanced, more accurate, or more valuable than it really is. It's the AI-era cousin of "greenwashing," where companies overstate environmental credentials to capture buyers who care about sustainability.
The FTC's enforcement initiative goes by the name Operation AI Comply, and it treats AI-washing as a straightforward violation of Section 5 of the FTC Act, the decades-old law banning "unfair or deceptive acts or practices." No new legislation was required. The agency's position is simple: if you claim your product does something using AI, you need evidence that it actually does — and that the claim would hold up to what a reasonable buyer understands it to mean.
The Case That Should Worry Every Small Business Owner
In May 2026, the FTC took action against CMG Media Corporation and two smaller marketing firms over an "Active Listening" tool they pitched to small business clients. The pitch: proprietary AI that could detect relevant conversations happening near a customer's smart devices — phones, smart speakers, connected TVs — and use that voice data, gathered with consent, to geographically target ads to people already interested in a business's products.
None of it was real. Investigators found the companies had no such listening technology. Instead, they purchased ordinary email lists from data brokers and resold them to small business customers at a steep markup, dressed up as cutting-edge AI targeting. Combined penalties across the case topped $930,000.
What makes this case notable isn't just the brazenness — it's who the victims were. Most AI-washing headlines focus on consumers being misled by chatbots or shopping tools. This case, and seven others like it, involved business-to-business deception: AI vendors lying not to end consumers, but to the small business owners buying their tools. The FTC has made clear it treats a small business defrauded by a vendor's false AI claims exactly like it would treat a misled consumer.
A Pattern, Not a One-Off
Once you look across the full run of Operation AI Comply cases, three recurring schemes emerge.
Fabricated wealth and performance claims. Air AI marketed "conversational AI" software as a drop-in replacement for human customer service reps, promising entrepreneurs and small business owners could build profitable businesses around it. The FTC alleged the company bilked buyers out of roughly $19 million in deceptive earnings claims. The settlement included an $18 million judgment (largely suspended based on inability to pay) and a permanent ban on marketing business opportunities. Similar cases hit Ascent Ecom, Ecommerce Empire Builders, FBA Machine, and Click Profit — the latter facing $20 million in judgments after investigators found roughly a fifth of customers earned nothing once fees were deducted.
Invented or wildly overstated capabilities. Workado sold AI-detection software claiming 98% accuracy at identifying AI-generated text. The FTC found the real accuracy rate was 53% — barely better than a coin flip. The resulting order barred the company from making accuracy claims it can't substantiate and imposed ongoing compliance monitoring.
AI tools built to help others deceive. Rytr, an AI writing assistant, was cited for marketing a feature specifically designed to generate fake testimonials and reviews — turning the tool into infrastructure for other companies' consumer deception.
Across all of it, the FTC's legal test doesn't change: what would a reasonable buyer understand the claim to mean, and can the seller back it up? Company size doesn't grant an exemption — the agency's 2024–2026 enforcement record includes vendors with fewer than 100 employees.
Why This Matters More for Small Businesses Than Big Ones
Large enterprises typically run new software vendors through procurement teams, legal review, and pilot programs before signing a contract. A five-person business rarely has that luxury. When a sales call promises an AI tool will "automate your bookkeeping," "screen resumes with zero bias," or "guarantee $10k/month in passive income," the pressure to trust the pitch — especially from a business owner already stretched thin — is real.
That asymmetry is exactly why regulators have started treating small businesses as a protected buyer class in these cases, not just as the marketers doing the misleading. It's also why the responsibility for catching a bad AI claim increasingly falls on the buyer, because contracts signed on faith are hard to unwind after the fact.
A Practical Vetting Checklist Before You Buy an "AI-Powered" Tool
Before signing up for the next AI-branded product pitched to your business, a few questions can save real money:
- Ask for substantiation, not marketing copy. If a vendor claims a specific accuracy rate, time savings, or revenue outcome, ask how it was measured and against what benchmark. A vendor unwilling to share methodology is a red flag.
- Get performance claims in writing, inside the contract. A verbal promise on a sales call is not enforceable. If "98% accuracy" or "automates 80% of your bookkeeping" is the reason you're buying, it belongs in the agreement — not just the pitch deck.
- Ask what "AI" actually means for this product. Is it a large language model, a rules-based system with an AI label bolted on, or something in between? Ask how the system was trained, whether it uses your data to train other customers' models, and what guardrails exist around automated decisions.
- Watch for "fully autonomous" language with no review step. Products that claim to run unsupervised — approving loans, screening candidates, managing ad spend — without a human-in-the-loop option or audit trail deserve extra scrutiny.
- Check for a real off-ramp. Confirm you can export your data and cancel cleanly if the tool underperforms. Vague answers about data ownership or deletion at cancellation are a warning sign.
- Verify independently before renewing. A tool that performed well in a sales demo should keep performing well eight months in. Revisit the original claims against real usage data before auto-renewal kicks in.
None of this requires a lawyer for every purchase — it just requires treating "AI-powered" as a claim to verify, not a feature to assume.
What Happens If You've Already Been Burned
If you've already bought a tool on the strength of an AI claim that turned out to be false, you're not out of options. Document everything — the original sales materials, emails, contract language, and any screenshots of marketing claims — before a vendor has a chance to quietly update its website. Small businesses can file a complaint directly with the FTC at ReportFraud.ftc.gov, and many state attorneys general run parallel consumer- and business-protection units that accept the same kind of complaint. If the purchase was made on a business credit card, a chargeback dispute citing misrepresentation is also worth pursuing — card networks have their own deadlines, so don't wait for a regulatory outcome before starting that process.
It's also worth checking whether the vendor is a repeat target. The FTC publishes its enforcement actions and consent orders publicly, and a quick search of a vendor's name alongside "FTC" or "AI-washing" before signing — not just after something goes wrong — takes a few minutes and can surface a settlement history that never comes up on a sales call.
The Regulatory Trend Isn't Slowing Down
Operation AI Comply survived a change in FTC leadership without losing momentum, and the agency has signaled the pace of enforcement will continue rather than taper off now that the initial wave of cases has established precedent. That's a meaningful signal for small business owners: the legal risk of relying on an unverified AI claim isn't a one-time news cycle, it's an ongoing enforcement posture. Vendors are on notice that "AI" in a product name or pitch deck now invites more scrutiny, not less — which, over time, should make substantiated claims easier to tell apart from hype. Until then, the burden of verification sits with the buyer.
Where This Connects to Your Bookkeeping
Vendor scrutiny matters just as much when the "AI" claim is attached to your financial tools. An AI bookkeeping assistant that promises to "auto-categorize 100% of transactions" or "eliminate reconciliation errors" is making a testable claim — one worth checking against your actual books before you rely on it for tax filings or investor reporting. This is one of the practical advantages of transparent, plain-text financial records: when a tool's AI-generated categorization or reconciliation looks off, you can open the underlying ledger file, see exactly what changed and why, and catch the discrepancy — instead of trusting a black-box dashboard that gives you no way to audit its own work.
Keep Your Finances Transparent and Auditable
As AI-branded tools multiply across every corner of small business software, the ability to independently verify what's actually happening in your books matters more, not less. Beancount.io provides plain-text accounting that gives you complete transparency and control over your financial data — no black boxes, no vendor lock-in, and every entry auditable in a format you own. Get started for free and see why developers and finance professionals are switching to plain-text accounting.