Most People Trust AI Less Than Humans. Startups Are Betting On That

The Federal Trade Commission has spent the past year building a real enforcement record against the data broker industry. Where the agency once settled for fines after the fact, it has shifted toward proactive actions that redraw the boundaries of what data brokers can collect and sell.

The FTC’s action against Gravy Analytics and its subsidiary Venntel demonstrates the change in tactic. The agency banned the company from collecting and selling location data precise enough to track individual visits to medical facilities, religious institutions and domestic violence shelters, in an unprecedented crackdown. A separate action against Mobilewalla marked the agency's first-ever prohibition on collecting consumer data from real-time ad-bidding exchanges. In February, it formally reminded data brokers of their obligations under the Protecting Americans' Data from Foreign Adversaries Act.

Regulators are catching up to a problem consumers are all too familiar with. More than half of consumers now trust AI less than they trust other humans with their personal data, up from 48% a year earlier, an increase that marks the largest single year-over-year shift in Usercentrics’ State of Digital Trust 2026 report. A separate global study from KPMG, surveying more than 48,000 people across 47 countries, found that while 66% of people already use AI regularly, only 46% are actually willing to trust it. And 70% say AI needs more regulation than it currently has.

A business built on the assumption that trust has to be earned back

Del Andujar founded TrueData Solutions on the bet that the trust issue is not a public relations problem AI companies can talk their way out of. "The AI economy is built on data, but the companies that ultimately win won't be the ones collecting the most information. They'll be the ones customers trust the most," Andujar says in an interview.

TrueData offers its core data broker opt-out tools for free, a structural choice that runs against the standard subscription model most privacy and data-removal services use. That subscription model exists for a structural reason. Most data-removal services run between $39 and $199 a year, with mid-range plans landing around $77 to $130 annually, according to category pricing comparisons. The recurring charge is not arbitrary. Brokers are permitted to re-collect information from public records as soon as 90 days after a removal, so most services rescan and resubmit requests every 60 to 90 days. Deletion is not a one-time event, which is exactly what makes it a durable subscription business.

"One of the biggest questions every AI company should ask itself is whether its business model aligns with the customer's best interests," Andujar says. "If solving a privacy problem reduces recurring revenue, there's an inherent tension that deserves scrutiny." A free core product invites the obvious follow-up question: how does the company sustain itself? TrueData is bootstrapped, with no outside investors pressing for the predictable recurring revenue that venture-backed privacy startups are built to produce. That removes one structural pressure toward subscriptions. It does not by itself explain what pays for the engineering behind a tool given away at no cost.

Set against the FTC’s recent enforcement record, that tension is not abstract. Several of the companies the agency has taken action against built entire business models on the same location and behavioral data TrueData helps consumers remove from broker databases in the first place.

Small businesses are collecting more customer data with fewer safeguards

Small businesses have adopted AI faster than they have written rules for using it, and that gap directly affects how customer data is handled. A small business using an AI chatbot, an email tool, or a scheduling assistant often feeds customer names, contact details, and purchase history into a third-party AI system with no formal review of where that data goes afterward. The vendor behind those AI tools could resemble the data brokers the FTC has spent the past year prosecuting.

KPMG’s research found that the majority of people believe AI needs stronger regulation, a warning to any small business assuming today's light-touch environment is permanent. That regulation is already arriving, and it is pointed directly at brokers. California's Delete Act required data brokers to register with the state's Privacy Protection Agency by January 31. As of August 1, registered brokers must check the state's Delete Request and Opt-out Platform at least every 45 days and honor deletion requests within 45 days of receiving them, according to the agency's system requirements. California has also stood up a data broker strike force to enforce it. The registration threshold turns on selling personal information about consumers a business has no direct relationship with. Most small businesses sit outside that line. Many of the vendors they hand customer data to do not.

Andujar sees the same pattern from the vendor side. "Privacy is no longer just a cybersecurity issue. It's becoming a business strategy," he says. “Companies that treat privacy as a feature will always struggle to differentiate themselves.”

What customers actually expect is narrower than ‘eliminate all risk’

KPMG’s global study found 54% of people are wary about trusting AI systems, and Usercentrics' data shows 52% of consumers say they will pay more for AI transparency.

Customers don’t expect zero risk. They expect honesty about the risk that exists. They are already acting on that expectation in ways that reach revenue. Nearly half of consumers, 47%, took at least one action in the past six months over how a company used their data in AI, according to the Usercentrics report, which was conducted by Sapio Research across 11,000 consumers in seven markets. Of those actions, 24% canceled a subscription, 20% switched to a competitor and 20% cut their spending. The consumers willing to pay more for transparency are paying an average premium of 7%. For a company with a million customers, that is a swing measured in six figures of purchasing decisions, not a sentiment score.

"Consumers don't necessarily expect companies to eliminate every privacy risk overnight, but they do expect transparency, honesty, and meaningful control over their own information," Andujar says.

What can a small business actually check right now?

Start with an inventory of every AI tool that touches customer data, not just the obvious ones like a chatbot. Take a look at scheduling tools, email marketing platforms, and CRM add-ons that quietly added AI features over the past year. For each one, find out whether customer data is used to train the vendor’s models, and whether that use is opt-out, opt-in, or not disclosed at all. A vendor that cannot answer that question directly is itself an answer, and the FTC's recent enforcement actions show regulators are now willing to treat "we didn't disclose it clearly" as a real violation.

Next, put a plain-language data statement somewhere a customer can actually find it before they hand over information, not buried in a privacy policy written for lawyers. This does not require legal language or a compliance department. This is simply telling a customer, in a sentence they can understand, what happens to their information and who else has access to it.

Finally, consider data-broker exposure as part of the same audit. Every AI tool now touching a small business's stack is a potential new path for customer or business data to end up aggregated and resold by a broker of the kind the FTC has spent this year prosecuting. Most small business owners have never checked whether their own information, let alone their customers', shows up in one.

"The future of AI isn't simply about building smarter algorithms," Andujar says. "It's about building organizations that deserve access to people's data in the first place."