Related articles
On-site photo
While AI has upended how information is created, shared, and trusted, it also holds the key to rebuilding that very trust, said Jany H.J. Zhao, founder of NextFin.AI and chairperson of TMTPost Group, at the NEX-T Summit 2025 hosted by NextFin.AI and Global Asian Leadership Alliance (GALA) in Silicon Valley.
AI itself may offer the answer — if used wisely. "AI has changed the entire information chain — and therefore the trust chain," she said. "It creates risk, but it's also the solution."
At a panel titled "AI, Media, and Public Trust," Zhao joined discussions with Adi Ignatius, the editor-in-chief of the Harvard Business Review; Dexter (Tiff) Roberts, the fellow of Atlantic Council; and Henny Sender, a senior advisor & consultant to BlackRock CEO.
Ignatius opened the discussion with a candid assessment of the shifting media landscape. "We're seeing lower barriers to entry and more viewpoints," he said, pointing to how social platforms have democratized publishing but also enabled the spread of misinformation and deepfakes. "There's been an erosion of trust that has accompanied that."
He noted that audiences are increasingly transferring their trust from institutions to individuals. "We're in an era where people follow personalities rather than publications," he said, citing historian Heather Cox Richardson's 2.5 million followers on the Substack platform as an example of this transformation.
But the same technologies that empower individuals also threaten established media models. Ignatius highlighted the unresolved issue of compensation: "AI has taken our content, built great platforms, and isn't remunerating us." The consequences are existential, he said. "We're in the late stages of the current model, and I haven't seen a publication that can survive the disruption that's about to hit us."
Roberts followed by citing sobering data from Pew Research Center. "Over half of Americans say AI in media is somewhat or very negative, and only 10 percent view it positively," he said. Even among those optimistic about AI's overall benefits for the U.S., most believe it's harmful to the media industry.
"The fear is all about misinformation and hallucinations," Roberts explained. "Two-thirds of people worry that what they're reading might be false." Despite that pessimism, he said, media organizations understand they must adopt AI or risk irrelevance.
He pointed to Bloomberg as a case study. The financial information giant has embraced AI to enhance its terminal functions, using machine learning for data extraction, summarization, and rapid information delivery. Bloomberg has even developed its own large language model, BloombergGPT, designed to improve financial analysis and journalism.
Still, Roberts emphasized that trust cannot be rebuilt without transparency and shared standards. "Media companies need to come together to set clear industry guidelines on AI use," he said. "And we must be open with audiences — tell them exactly where and how AI is being used, and ensure that human journalists are still in the loop for fact-checking and oversight."
Sender, also the former chief correspondent of Financial Times, offered a sobering reflection on journalism's identity crisis in a polarized era. "In a fragmented world, how can media not reflect that fragmentation?" she asked. "We used to define our role as speaking truth to power, but now there's no consensus on what truth is."
She warned that the pursuit of "balance" often leaves newsrooms vulnerable to manipulation. "If someone claims Obama wasn't a U.S. citizen and we report it, people might believe it's true. But if we add context saying there's no evidence, we're accused of editorializing. We can't stay separate from the world anymore."
The result, she said, is a perception that journalists are advocates rather than educators. "That's my downbeat message — the media is trapped between its mission to inform and an audience that no longer agrees on reality."
Zhao, also the publisher of Barron』s China, described how her company NextFin.AI develops AI agents that "filter out noise and leave only the signal," a critical capability for financial media. "AI can cross-check information sources, compare models, and verify facts faster than any human team," she said. "That's how we can rebuild trust."
She argued that trust is inseparable from sustainability: "Without trust, no media business — AI or traditional — can survive."
Zhao also offered a pragmatic view on the evolving newsroom. "In the past, Bloomberg needed thousands of employees to gather information globally," she said. "With AI, we can achieve similar depth with 10% of that workforce — not by replacing people, but by making them more efficient." The key, she added, is to invest in what consumers truly value: "fast, deep, and transparent information."
Returning to the conversation, Ignatius emphasized that legacy media's financial model is breaking down. "In the U.S., many publications have become philanthropic ventures," he said. "A wealthy individual buys them, sustains them until they get tired of it."
For long-term survival, he argued, media must pursue a dual bottom line — financial viability and social responsibility. "You have to believe in both missions," Ignatius said. "Be creative with revenue, interact more closely with readers."
He envisioned a future where AI enables highly personalized engagement. "At theHarvard Business Review, maybe our model could be guiding subscribers through a journey — knowing who they are, what they aspire to be, and helping them get there," he said. "Content in the future will be for an audience of one."
Roberts reflected on how AI is transforming editorial decision-making. "When I was Asia news editor for Bloomberg Businessweek, we'd have a weekly call deciding what stories readers needed to know," he said. "That model already feels obsolete."
AI-driven personalization, he said, allows each reader to receive customized news. "That's powerful, but it raises new questions: what happens to the watchdog role of the press? What about investigative journalism that exposes wrongdoing people didn't even know existed?"
He warned that hyper-personalization could deepen social divisions. "If everyone only sees the news they want, we risk creating AI-powered echo chambers," Roberts cautioned. "That's a real danger — and the one the industry hasn't yet addressed."
Sender lightened the mood with humor: "Netflix thinks I'm a pathological murderer," she quipped, illustrating AI's imperfections. But her point was serious: "AI is flawed, and the media's economic incentives often blur the line between information and entertainment."
She called for systemic reform of incentive structures that prioritize engagement over truth. "We're economic animals," she said. "We sell attention. The conflict between headlines and content has always existed — AI just amplifies it."
Zhao added that media should be strategic partners for companies and communities. "Entrepreneurship is about solving problems," she said. "Media should help identify and solve consumers' real problems — not just broadcast information."
Ignatius agreed, offering an example from a venture he's involved in that pairs startup founders with journalists to refine their storytelling. "Many founders are brilliant at math and code but terrible at telling their story," he said. "Helping them communicate isn't lowering standards — it's empowering innovation."
He called for openness to new collaborations, saying that media institutions can share voice with responsible sponsors without compromising integrity. With the right guardrails, partnerships can enhance value, sustain operations, and even strengthen trust.
