The single most essential ingredient of good journalism is trust. By that measure, journalism is in a very bad way: According to The Reuters Institute’s annual survey, only 37 percent of the global public trusts the news industry, down nearly 20 points from ten years ago. In the US, that figure stands at 25 percent.
Reuters also reported that for the first time, more than half of respondents get their news primarily from algorithmically driven platforms like Instagram, TikTok, and YouTube. And the fastest growing new “surface” for news consumption? AI chatbots like OpenAI’s ChatGPT and Anthropic’s Claude.
The central story of journalism in the digital age is one of collapse. Platforms and their algorithms have cleaved audience from traditional news brands. The industrial-scale AI that runs sites like Instagram and YouTube are tuned for engagement and advertising sales, not trust or truth. Personalized AI of chatbots like Gemini or ChatGPT seem destined for a similar fate.
Against this cheery backdrop, a glimmer of hope: This morning NewsGuard*, which has been scoring the trustworthiness of news sites for nearly a decade, announced NewsGuardAI, a new service that packages AI chat with the company’s extensive database of trusted sources and narratives.
“NewsGuard AI combines our proprietary database of 12,000 reliable publishers, vetted for reliability by our analysts, with a custom-built AI process that instantly delivers accurate news aggregated from quality sources,” NewsGuard said in its release announcing NewsGuardAI. “NewsGuard journalists trained NewsGuard AI with 41 editorial safeguards designed to provide responses that cite sources for every fact, report all sides of a story even-handedly with nuance, and suggest follow-up prompts based on a reader’s interests.”
I’ve tried NewsGuardAI, and while it’s early, I find it super useful. I asked it for a summary of this morning’s news, and after seeing mention of the hotly contested Democratic primary race in New York’s 12th Congressional District, I used it to dig into details (fittingly, the core issue is AI regulation, and the AI industry is spending a ridiculous amount of money to spike the campaign of a pro-regulation candidate). As you can see from the screen shot below, every assertion made by NewsGuardAI is backed by citations from reliable sources. 
NewsGuard also announced a business model for its new AI offering. While initially a free service, the company will charge a $6 per month subscription after a threshold of queries is met. Revenues will be split with publishers whose stories are cited.
The question then becomes: Will anyone actually pay for trusted news, even packaged in the shiny new clothes of an AI chatbot? And in a world dominated by social media and video platforms, how will anyone even find NewsGuardAI in the first place? The company has considered this question, and answers it with a distribution strategy based on partnerships with its trusted network of publishers, including The Atlantic and the fact checking site Snopes.
I’m encouraged by the emergence of services like NewsGuardAI, but it faces an uphill battle. The service is text heavy and presumes a level of engagement with the news that most consumers no longer possess. That said, if NewsGuardAI can become a single source of trusted news for those who drive the news – journalists, researchers, and yes, even the “news influencers” on YouTube and TikTok – it may well offer a foundation of truth and trust for an ecosystem that’s proven hostile to facts. And that feels like a win to me.
*I was an early advisor to NewsGuard when it launched, and as such hold advisory shares in the company.
–