AI Trust Crisis: Why Users Draw the Line at News and Politics

While AI adoption surges across everyday tasks, a clear boundary emerges for serious information. Insights from the Social Lens Library reveal users across all generations share unified skepticism: AI isn’t ready for news and political content.

The Trust Divide

AI has found widespread acceptance for productivity tasks, list-making, process optimization, and routine searches. But mention politics or current events, and users pump the brakes. As one Asian Gen Z user explained: “I do use ChatGPT, but I do not use AI tools at all for political information. Just because I’m not sure that I trust ChatGPT for more serious topics. I trust it to optimize processes, to make lists, make plans, regular searches, but for things that include opinions and facts in terms of real-world issues, I’m not going to trust ChatGPT with that.”

This sentiment spans demographics. A Millennial Latino echoed similar concerns: “I don’t use AI for political information. I don’t trust it. These systems are fed specific data, and their algorithms aren’t designed to be genuinely helpful. AI has no real understanding of complex issues. Just look at how Grok was manipulated into extreme positions. That actually happened, and it’s alarming. I can’t believe people would trust AI for news.”

Core Concerns Drive Hesitation

The resistance stems from three primary worries: potential bias from AI creators, the risk of false information, and the absence of reliable filtering mechanisms. Users recognize AI’s vulnerability to manipulation, as one White Gen X participant noted: “To me, AI corrupts our opinions. I don’t trust AI answers 100%. I use it for building ads, scripts, resumes, and pictures, but you can really manipulate the truth and make it go any way you want.”

The desire for authentic information remains paramount. A Gen X Black user emphasized, “I don’t use AI for political news or information. Not ChatGPT, browser extensions, or any of those tools for politics because they’re too easily manipulated. I want real news from real sources, especially when dealing with something as important and delicate as politics.”

The Source Problem

Perhaps most telling is users’ frustration with AI’s source selection. A Latina Millennial highlighted a critical flaw: “I absolutely don’t use AI for political information. It can’t be trusted at all. Even when you tell it to use credible sources, it thinks non-credible sources are credible. It uses Reddit constantly, and that’s just people’s opinions. Reddit can have good information and people are sometimes right, but for politics you need facts, not opinions.”

The Path Forward

As AI companies face growing scrutiny, addressing these trust gaps becomes essential. Users are calling for source transparency, improved filtering of unreliable content, and clear quality indicators.

Meanwhile, political candidates must adapt to an AI-driven information landscape. Key strategies include strengthening digital presence with consistent, updated messaging across platforms, using clear language that resists distortion, and monitoring how AI represents their positions. Candidates should optimize for discoverability through structured Q&As and official content, build transparency about their own AI use and fact-checking processes, and focus on authentic, verified information that builds trust across generations.

Political campaigns can lean into AI to get a head start on making their content more accessible, trustworthy and ready for AI. This is good politics and good for trust.

Visit our Social Lens Library Stories Playlist, Human Stories @Scale on YouTube here.

Note: Our team uses AI to generate voices in videos (protecting participant identity), edit content, and analyze video transcripts. Human oversight guides all aspects of our work.