Copy article

AI is making online advertising’s trust gap worse: transparency needs to become measurable, not voluntary

ended 20. March 2026

Online advertising has always had a trust problem. Fraudsters move faster than standards, and the supply chain is designed to hide who made money from a bad outcome. Now add AI-generated creative, AI-targeting, and automated optimisation, and the system becomes even harder to audit.

A government-backed taskforce is a sensible response, but the question is what kind of progress is being measured. Creating working groups and voluntary principles is easy. Proving that ads are safer, cleaner, and more accountable is much harder.

The uncomfortable truth is that AI makes both sides better. It helps brands spot waste and improve targeting. It also helps criminals scale up impersonation, deepfakes, and fake storefronts. The public mostly experiences this as background noise until it becomes financial harm.

The next phase needs less talk about innovation and more talk about evidence. Which platforms can trace an ad back to a verified payer? Which intermediaries can prove they screened out obvious fraud? And which parts of the chain still operate on “trust us” data?

Questions:

  • Should ad platforms be required to publish fraud rates and enforcement metrics?
  • Where should liability sit when AI-generated scams are delivered through paid ads?
  • What would meaningful transparency look like across the ad supply chain?
  • Can voluntary standards keep up, or is tougher regulation now inevitable?
  • How should small businesses protect themselves from AI-enabled advertising fraud?

3 responses from the Newspage community

Copy all

Star Quote
Copy

Rohit Parmar-Mistry
Founder at Pattrn Data
Voluntary principles are fine, but they fail the moment money gets tight. If platforms want trust, publish numbers that can be compared: confirmed fraud rate (with a definition), median time to takedown for scam ads, and the share of spend from verified payers before an ad runs. Then make the evidence auditable by independent reviewers, not just "trust us" dashboards.

On liability, treat paid distribution like a safety-critical system. Whoever can prevent harm cheapest should carry the obligation. In practice that is the platform and the biggest intermediaries, because they see the signals first and still take the fee when things go wrong.

AI will keep raising both offence and defence. Regulation should focus on controls you can test: identity, provenance of creatives, and enforcement metrics.

For small businesses, assume you will be spoofed. Lock down your domain and email auth (DMARC), monitor brand keywords, and use verified advertiser programmes where available.
Copy

Editor
Mitali Deypurkaystha
AI Governance Strategist & CMO at AISensa
A £10 billion ad market now runs on “AI-fuelled fraud factories” where platforms profit from volume while victims pay the price. The uncomfortable reality is that AI hasn’t just widened the trust gap, it’s weaponised it, making scams, impersonation and fake ads faster to produce and harder to trace. I saw this first-hand when a friend spent months trying to remove a deeply personal video uploaded without her consent, while the platform continued to benefit from the engagement it generated. That same “content is content, spend is spend” mindset also applies to advertising, where a pound from a scammer is treated no differently to a legitimate business. Until platforms are forced to prove where money comes from and take real responsibility for harm, with huge ramifications for them if they don't, trust will remain a slogan, not a standard.
Copy

Colette Mason
AI Ethics Consultant at Clever Clogs AI
Liability should sit where the money lands. If a platform takes payment to deliver an ad, it owns the outcome of what it delivered. That's not radical. It's how every other paid service works. You don't get to collect the fee and then claim you were just the postman when the parcel explodes.

The reason this hasn't happened is because the ad supply chain was built to make accountability structurally impossible. Every intermediary points at the next one. AI just accelerates the speed at which harm travels through that chain while nobody admits they touched it.

As for small businesses protecting themselves? Honestly, they can't. Not individually. A sole trader doesn't have the tools to audit programmatic ad placement or detect a deepfake storefront cloning their brand. That's precisely why this can't stay voluntary. When the only people who can afford protection are the ones who least need it, you don't have a standard. You have a ladder to pull up leaving the others to flounder.