Building trust in the age of AI: from the newsroom to the niche

Part 1: Inside the FT’s AI newsroom – where trust & transparency shape the product

“Time and trust – those are the two currencies we keep exchanging between individuals and businesses,” says Peltea, opening her conversation with Liz Lohn, director of product, AI and editorial tech at the Financial Times. For Lohn, that exchange is playing out daily in how the FT experiments with AI while protecting its most valuable asset: reader trust.

Rather than waiting for the hype to pass, the FT leaned in early. “The company invested in having a product team that is focused on AI solely for some time,” says Lohn, who leads that team from the side of product and technology. “It meant we could look at opportunities and threats, what competitors were doing, without being bogged down by backlogs.”

The result? Two strategic tracks. One supports journalists in creating new stories through investigative analysis of vast datasets. The other focuses on augmenting the production process – editing, error-checking, SEO and content transformation for different formats. But never fully replacing the human writer. “Our original reporting is not going to be written by AI. It will always be written by humans,” she says.

This clarity extends to how AI-generated text in newsroom experiments is communicated to readers. “We’re using labeling if it is drafted by AI… [like] ‘AI-drafted, human-reviewed,’” Lohn explains. The FT even tested features like AI-drafted article summaries in an A/B test to gauge impact: “We looked at several engagement metrics including the scroll depth and they either slightly increased or stayed within ‘do no harm’ decrease. Most importantly, there were no factual edits to summaries which, coming back to the trust factor, gave us confidence to continue with this feature.”

In fact, trust became a surprising source of innovation. As Lohn reveals, “We asked [readers] what they do with our content using AI – and they told us explicitly they copy-paste our articles into ChatGPT to get a summary.” Instead of resisting, the FT integrated that into the user experience.

This approach isn’t just about tech, it’s about listening. “We talk to people. We look at industry statistics. People do want transparency when AI is used,” she says. And when it comes to personalization, the FT is carefully walking a fine line: “Users want to know the data that went into their recommendations… When a recommendation is bad, they want to fix it; and when it’s really good, they feel like it’s creepy and maybe the model knows too much.”

It all points to a new kind of product strategy – one that prioritizes content created by human experts, audience segmentation, and transparency. As Lohn puts it: “People come to us for trusted humans who write journalism, but they also understand the convenience factor and the user experience.”

Part 2: Fashioning trust in a fragmented world – lessons from Enstyle

From premium journalism to curated fashion, the question of trust isn’t just about content, it’s about consistency. That was the theme running through the second conversation between Usercentrics’ Peltea and Alina Veselaya, chief marketing officer of fashion trend intelligence platform Enstyle.

“When it comes to the definition of marketing and what its primary goals are: one is be found, and the second call to action,” says Veselaya. “There is a lot of stuff happening between these two and this is where the question of trust comes in.”

For Veselaya, building that trust across a non-linear customer journey means maintaining a coherent brand identity, even across increasingly algorithm-driven environments. “You have to think about how consistent it is across channels, so that when the person comes across your brand again and again, they can recognize it – and form those associations that drive trust.”

In an age of performance metrics and click obsession, Peltea sees a resurgence in brand-first thinking. “There’s the rise of the brand again,” she says. “People trust the company if they can see what it truly is… I trust you with my data because I know your practices, and I trust that you represent something.”

That notion of brand identity takes on new dimensions when applied to personalization. Enstyle is seeing a cultural shift from mass trends to “super niches.” TikTok, Veselaya notes, has fueled the rise of micro-communities like ‘grunge-amour’ and ‘loser-core.’ “You as a brand can decide which micro-communities you relate to and want to have a conversation with.”

But trust isn’t just about what you stand for, it’s about how you show up – especially in those first few digital moments. “First impressions can never fully represent a brand,” says Veselaya, “but it can be enough for the person looking at that website to want to know more about it.”

That’s where Usercentrics’ First Impressions Matter campaign lands its point – reminding marketers that consent banners and privacy settings aren’t just compliance checkboxes, but part of the brand experience. As Veselaya adds: “People right now… they’re not always time-generous, so it’s really important for brands to respect that – and the idea of saving their time.”

A shared message: trust is the strategy

Across both conversations, one thread is clear: trust isn’t a soft metric, it’s a strategic foundation – built through transparency, consistency and control. Whether you’re refining bullet-point summaries at the Financial Times or targeting niche tribes on TikTok, trust isn’t a byproduct of good marketing. It is the marketing.

As Peltea reflected in both discussions, “We’re moving away from all upfront data asks, and being more mindful about the value we create at every touchpoint.” And that’s what makes Privacy-led Marketing so powerful – not because it hides complexity, but because it helps simplify the experience, on the user’s terms.