The Great Trust Recession

I spent twenty years inside the systems that eroded trust online. That’s why I’m building the one designed to restore it.

There was a kid at our LEGO Star Wars launch event. Maybe seven years old. Homemade Darth Vader costume, the kind where you can see the tape holding the cape to the collar and the helmet is a little too big and keeps sliding. His mom was standing behind him looking tired and proud in equal measure. He’d made her bring him. He wasn’t there for the free stuff or the screens or the demos. He was there because Star Wars was part of who he was, and this was the room where it was happening, and he needed to be in it.

I think about that kid a lot. And it’s not nostalgia. It’s that he’s the clearest picture I have of what trust looks like before anyone figures out how to monetize it. A seven-year-old in a taped-together costume who feels like he belongs to something bigger than himself. He didn’t need an algorithm to find that event. He didn’t need a notification to care. The product and the community around it had earned something that can’t be manufactured: his genuine, unguarded belief that this thing was his.

I spent the next twenty years trying to build that feeling in every product, every community, every company I worked with. I got good at it. At AdRoll, at Sojern, at Relativity Space, at YouCaring, I helped build brands that people genuinely connected with, that earned real trust, that mattered to the people who used them.

But I was doing it inside systems designed to do something else entirely. Systems that measured trust as a retention metric, loyalty as a conversion rate, and belonging as time-on-site. I was building genuine community with one hand while operating inside an architecture that was systematically eroding trust with the other. The kid in the Darth Vader costume became a “monthly active user.” His devotion became a data point. And the platforms that were supposed to connect people like him started optimizing for engagement at the expense of everything that engagement was supposed to serve.

48% of people now question the authenticity of almost everything they see online. Researchers are calling it The Great Trust Recession. When I read that stat, I didn’t feel surprised. I felt responsible. I’ve spent twenty years inside the systems that produced this outcome. I understand exactly how we got here. And that’s why I’m building the alternative.

How Bad Is the Trust Crisis Right Now?

Researchers are calling it The Great Trust Recession. I think the name undersells it. A 2026 survey of 2,000 people in the US and UK found that 48% now question the authenticity of almost everything they encounter online. Not some things. Almost everything. 43% say they no longer trust most online content. Over half have changed their online behavior entirely, retreating to a small number of trusted sources and ignoring the rest.

This isn’t abstract anxiety. It’s showing up in concrete ways. Consumer fraud losses hit $12.5 billion in 2024. Deepfake incidents increased 257% from 2023 to 2024. AI-generated “ghost businesses” are appearing online, entities that look legitimate but have no real-world presence, and INTERPOL and the FTC have both issued warnings about the sophistication of AI-enabled fraud. Nearly half of consumers now prefer brands that avoid using generative AI in customer-facing content altogether.

That last one stopped me. Consumers actively prefer brands that don’t use AI to talk to them. We’ve built the most powerful content generation tools in history and the market response is: please stop using them on us.

How Did We Get Here?

The trust collapse didn’t start with AI. AI accelerated something that was already well underway. The foundation was laid over fifteen years of ad-driven platform economics. The mechanics are straightforward once you see them, and I saw them up close for a long time.

An ad-supported platform makes money when people spend time and interact with content. The algorithm maximizes those interactions. It can’t tell the difference between an interaction that enriched your afternoon and one that left you anxious and misinformed. A rage-click and a genuine connection look identical in the data. Engagement is engagement. I remember sitting in a meeting where we were celebrating a campaign’s engagement numbers and I realized that half the engagement was people arguing in the comments. By the metrics, it was our best-performing content that quarter. By any human measure, it was making people’s days worse.

Over time, this creates a bias toward content that provokes reaction over content that builds understanding. Sensationalism outperforms nuance. Outrage outperforms thoughtfulness. Volume outperforms craft. The platforms didn’t decide to degrade trust. The incentive structure did it automatically, the way water finds the lowest point. I watched it happen from inside companies full of good people building genuinely useful products, and I couldn’t have told you the exact moment when the system started doing more harm than good. That’s the thing about structural problems. They don’t announce themselves.

Why Is AI Making It Worse?

AI didn’t break trust. But it took a system that was already producing distrust at scale and removed the last constraint: the cost of making content.

Generative AI collapsed the cost of content creation to near zero. That means the same incentive structure that rewarded volume over quality now operates in a world where volume is effectively infinite. Feeds that were already noisy became unusable. Content that was already hard to verify became impossible to verify. The signal-to-noise ratio didn’t just decline. It inverted.

56% of users report seeing AI-generated content they consider low quality on their feeds often or very often. 66% say they’re more selective about what they engage with than a year ago. I talk to people who’ve essentially stopped trusting their own feeds. They’re not leaving platforms because they’re bored. They’re leaving because they genuinely cannot tell what’s real anymore.

And the platforms can’t fix it. That’s the part that most people miss. An ad-supported platform profits from content volume regardless of its quality or authenticity. Every piece of AI-generated content, real or fake, helpful or garbage, generates impressions that generate ad revenue. The business model is structurally misaligned with the user’s need for trustworthy information. You can’t patch that with better content moderation. The incentive layer is broken.

Why Can’t We Just Fix the Existing Platforms?

The obvious question is why the existing platforms can’t just fix this. Add trust features. Better moderation. AI labels.

They can’t, and the reason is structural, not operational. Trust isn’t a feature. It’s an architecture.

You can’t bolt trust onto a system that was designed to maximize engagement. Adding a blue checkmark doesn’t make a platform trustworthy. Adding “AI-generated” labels doesn’t fix the flood. These are cosmetic interventions on a structural problem, like putting a fresh coat of paint on a building with a cracked foundation.

Trust requires three things that existing platforms weren’t built to provide. First, identity verification that’s meaningful, not performative. Not a blue badge you can buy for $8 a month. Actual verification that the person behind the account is a real human with behavioral integrity. Second, discovery that’s driven by genuine relevance, not engagement optimization. An algorithm that shows you what’s good for you, not what’s most likely to keep you scrolling. Third, a business model that gets healthier when the user experience gets healthier. Not one that profits from outrage, addiction, and volume.

No existing platform has all three. Most don’t have any. And you can’t retrofit them because the business model, the ad model, actively resists all three. A platform that reduced engagement in favor of quality would see its revenue decline. Nobody’s going to do that voluntarily.

What Would a Trust-First Platform Actually Look Like?

I’ve spent years working through what a trust-first platform would actually require. Not from theory. From twenty years of watching what works and what doesn’t, up close, across very different industries.

You have to know who’s real. I learned this producing events for twenty years. Every person in those rooms was physically present. There was no question about whether they were real, whether they were who they said they were, whether they actually cared about being there. Online, we’ve lost that entirely. A platform that can’t verify that the person behind an account is an actual human isn’t a social platform. It’s a stage for bots and impersonators. Identity verification isn’t a feature. It’s the foundation.

Discovery has to serve the person, not the ad buyer. In gaming, people found the communities they loved through genuine interest. They showed up at that LEGO Star Wars event because Star Wars meant something to them, not because an algorithm decided to surface it. The feed model broke this by optimizing for engagement, which rewards whatever provokes the strongest reaction. Discovery built around genuine interest produces a completely different experience. You find things that matter to you instead of things designed to exploit your attention.

The people who build the community have to see their contribution reflected back to them. At my best events, the regulars were the ones who made the room work. They brought the right people, they set the tone, they kept coming back. And they knew they mattered. On current platforms, the person who helps fifty newcomers and the person who posts rage-bait get measured by the same engagement metrics. When people feel like stakeholders instead of content, they invest differently. I’ve seen it happen in physical rooms for two decades.

And the business model can’t be ads. Full stop. At YouCaring, we proved this. No platform fee. Trust-based tipping. Over 70% of donors tipped voluntarily. The business got healthier when the user experience got healthier. That’s the alignment that ad-supported platforms can never achieve, because their revenue depends on volume and engagement, not quality and trust. Revenue has to come from access, participation, and value exchange between creators and their communities. The business model and the user’s wellbeing have to pull in the same direction.

Why Am I Building This?

Because I’ve seen trust work as a business model, and I’ve seen engagement work as a business model, and they produce completely different outcomes for the people inside them.

The gaming communities I helped build around Star Wars and Nintendo lasted because the products earned genuine devotion. The events I produced in San Francisco for twenty years lasted because the rooms earned people’s trust and presence. YouCaring’s trust-based tipping model generated over $1 billion in donations and produced the largest crowdfunding campaign in history. Every one of these operated on the same principle: when you align the economics with the experience, people invest in ways that compound. When you misalign them, people disengage in ways that compound too. We’re watching the second version play out across the entire social internet right now.

Wondr is the convergence of everything I’ve built across gaming, events, ad tech, crowdfunding, and aerospace. Identity that’s verified. Discovery that’s genuine. Participation that’s rewarded. Economics that don’t require exploiting the people who create the value. I’m not building it despite my career inside the attention economy. I’m building it because of it.

What Happens If Nobody Fixes This?

The Great Trust Recession won’t self-correct. The forces driving it, AI content generation, ad-model incentives, deepfake proliferation, are all accelerating. Without structural intervention, we’re heading toward a digital environment where most people trust almost nothing they see online, where creators can’t build sustainable careers because their audiences don’t trust the platforms, and where the internet’s original promise, connection, community, shared knowledge, degrades into noise.

Better moderation won’t fix it. Better labels won’t fix it. Better intentions definitely won’t fix it. The fix requires building an alternative from the ground up, with trust baked into the foundation instead of painted on afterward.

Somewhere right now there’s a seven-year-old who loves something the way that kid loved Star Wars. Loves it enough to make a costume out of tape and cardboard and make their mom drive them two hours to another town. The question is whether the platforms that kid grows up on will honor that trust or exploit it. I spent twenty years inside systems that chose exploitation, usually without meaning to. I’m spending the next twenty building the one that doesn’t.

•  •  •

If this resonated:

For investors: Wondr is raising a founding round to build the trust-first social platform. If the thesis interests you, I’d welcome the conversation.

For creators and community builders: If you’re tired of building on platforms that don’t serve you, we’re looking for founding creators who want to help shape what comes next.

For everyone: Follow along. The trust crisis affects everyone online, and the solutions will too. Subscribe for essays on ownership, community, trust, and the future of social.

•  •  •

About the Author

Maly Ly is the Founder & CEO of Wondr, an AI-native social and discovery platform, and the founder of a growth lab advising early-stage startups. She is a founder, growth executive, and operator who has helped scale multiple startups to breakout growth and unicorn status across AI, Web3, aerospace, SaaS, and consumer tech.

Her experience includes leadership roles at category-defining companies such as AdRoll—named the Inc. 500’s #1 Fastest Growing Marketing Company—and Relativity Space, which reached a $2.3 billion valuation and became the second most valuable private space company after SpaceX. She has also held leadership positions at Eventbrite, Sojern, YouCaring (later acquired by GoFundMe), and SecurityPal AI, and earlier in her career helped launch top-selling products for franchises including Star Wars, Tomb Raider, and Nintendo.

Beyond tech, Maly spent two decades producing art, music, and  corporate and underground events, while leading digital campaigns for global brands including Mercedes-Benz, Aston Martin, Burberry, and Williams Sonoma. She is also a Certified High Performance Coach.

Her work has been recognized by Forbes, Fast Company, and Direct Marketing News with its Hall of Femme honor. She and her work have been featured in The New York Times, The Washington Post, CNN, NPR, Forbes, The Tonight Show, and The Ellen Show.