The Trust Arbitrage: Why Gen AI’s Advertising Ambition is a High-Stakes Gamble

Elery Pfeffer, the CEO and founder of Nift, points to a pivotal moment in the history of Silicon Valley, less than two years ago, when Sam Altman, the face of OpenAI, described advertising as a “last resort” for ChatGPT. At the time, the statement felt like a moral line in the sand. But today, that line has been crossed.

As OpenAI rolls out its advertising platform, the industry is watching more than just a new revenue stream. They are watching a real-time stress test of the most fragile asset in the digital age: user trust.

“It’s really important to understand how trust comes into play with the user experience,” Pfeffer says. “When you talk about social media, that’s an entertainment product. Entertainment has been interrupted by ads since the first newspapers. But a chatbot is different. We use chatbots to make decisions. The minute an experience is perceived to be self-interested, it erodes trust.”

The “last resort” moment

For Pfeffer, the shift at OpenAI isn’t a surprise, but it is a sobering reminder of the realities of scale. ChatGPT serves hundreds of millions of weekly users and processes billions of prompts daily. The capital expenditure required to maintain that infrastructure is staggering.

“This is the ‘last resort’ moment,” Pfeffer says. “Monetization was inevitable because you have to create a sustainable business, but they wouldn’t have done it unless they had to. Now, the experience becomes tainted in some respect, especially regarding trust.”

The risk, according to Pfeffer, is what he calls “trust arbitrage.” Unlike a search engine, where the user is presented with a list of links and must do the work of vetting the source, an AI agent distills information. It provides a singular, authoritative answer.

“With an AI agent, you as a user are pretty far away from the raw data,” Pfeffer says. “You’re taking a leap of faith that the chatbot distills information in a way that serves your interest and not an advertiser’s. If users begin questioning whether a response is genuinely helpful or commercially motivated, that behavioral shift is almost impossible to reverse.”

A PR battle for the mind

OpenAI’s rivals are already capitalizing on this vulnerability. Anthropic recently signaled its skepticism through a high-profile campaign that humorously questioned whether ads risk turning helpful assistants into biased salespeople.

“Anthropic was very clever,” Pfeffer says. “They created the fear in users that the information they’re getting is self-interested. OpenAI is now playing a losing PR game where they have to convince users that ‘the best answer always comes before the paid answer.’ But once you’re in the mindset of questioning the platform’s motives, the damage is done.”

Pfeffer likens the situation to the rise and fall of other “arbitrage” businesses. He points to the cautionary tale of WeWork, which signed long-term leases while offering users flexible usage. “OpenAI is building computing infrastructure for 2027 and 2028, betting on usage. But usage can disappear immediately if the trust disappears.”

From people to outcomes: the Nift philosophy

The tension between monetization and user experience is something Pfeffer has spent years solving at Nift. Nift’s core model — matching consumers with new products and services they can try for the first time — was built on a different premise: incentive alignment.

“The journey in starting Nift was really interesting when it comes to AI,” Pfeffer says. “I realized that when you have a machine making decisions, it’s better to sell outcomes instead of software. When we started Nift, it wasn’t about selling a ‘seat’ to a marketer; it was about the outcome of a consumer discovering something they actually love.”

This shift from “per-seat” (or per user) SaaS models to “per-outcome” models is, in Pfeffer’s view, the future of the entire industry. “A legal software sold by the seat is being put at risk by AI engines that sell the outcome of a legal document. The closer you can align with the value generated, the better you’re going to be.”

The intermediation of the brand

One of the most profound shifts Pfeffer has observed in his conversations with major brands — from fintech giants to retailers — is the loss of control. Though brands have been accustomed to tightly controlling their messaging, in the AI world the chatbot is the gatekeeper.

“Brands are getting disintermediated,” Pfeffer says. “You can now go on a chatbot and make a purchase directly. All the work that went into product marketing, creative, and on-site experience gets collapsed into a third party the brand has no control over. This puts the chatbots in the driver’s seat of the destiny of the brands.”

However, Pfeffer isn’t entirely cynical about the future of advertising. He believes there is a way to “thread the needle” if platforms look toward integration rather than interruption.

“I think Uber is doing a great job,” he says. “They have a juggernaut of a media business, but they protect the consumer experience by integrating ads in a way that feels natural to the journey. I recently flew United to Australia, and they had an integrated experience where picking a movie involved an ad. It was clearly an ad, but it wasn’t interruptive. It didn’t hurt the experience.”

The road to Cannes

As the industry prepares for major milestones like the Cannes Lions International Festival of Creativity, the conversation around AI advertising is expected to shift from “if” to “how.”

“By the time we get to Cannes, if OpenAI has released its platform, the conversation will move from theoretical fears into practical, operational questions,” Pfeffer says. “How much traffic can you get? How well does it convert? How do you advertise within the tiny bandwidth you have with a consumer on a mobile screen?”

Ultimately, Pfeffer believes the winners won’t be the platforms with the most data, but the ones that can maintain the “leap of faith” their users take every time they type a prompt.

“It becomes a battle of how self-interested the experience feels,” Pfeffer says. “If you create a seamless experience where the consumer is opting in and knowing what they’re getting, it can be a positive experience. If you erode that trust, you end up with less revenue than where you started.”

Arthur Zaczkiewicz is a former executive editor at WWD and has appeared in WWD, WWD Voices, Beauty Inc, Footwear News, WWD JAPAN.com, Yahoo Entertainment, Yahoo Life, Yahoo Tech, MSN, Yahoo Canada, Yahoo News, Yahoo Singapore, Aol, Los Angeles Times, Yahoo Finance, Flipboard, Yahoo Sports, NewsBreak, Sourcing Journal, and Athletech News.