Campbell Brown, former head of news at Meta, states that platforms optimizing for engagement will always prioritize hyperbolic content over accuracy. This creates a core, unresolvable conflict that undermines sustainable partnerships with high-quality news publishers, a problem AI's enterprise model might solve.
Large language models present all information, including falsehoods, with a consistently confident and articulate style. This fluency builds a quiet trust in the user, making them more susceptible to believing dangerously wrong information, particularly on high-stakes topics like health and politics.
The financial relationship between AI companies and news publishers will likely be settled by courts, not boardrooms. Campbell Brown predicts that ongoing litigation, like the New York Times vs. OpenAI lawsuit, will be the forcing function that establishes a business model for content ingestion, rather than proactive deal-making.
AI models can synthesize information and write more effectively than many generalist reporters. Campbell Brown argues this shift makes deep, nuanced expertise and original reporting the only defensible value for journalists. The profession is moving away from generalism towards highly specialized, entrepreneurial creators.
To evaluate AI responses on complex topics like politics, expertise alone is insufficient. Campbell Brown’s company seeks experts like former CIA analysts who are trained to remove personal bias, consider all possibilities, and focus on the correct *framework* for an answer, rather than a single 'right' one.
With trust in media institutions at an all-time low, consumers are increasingly getting their news from individual creators they trust, such as podcasters and newsletter writers. Campbell Brown notes this is a fundamental consumption shift, where the personal brand of a creator replaces the institutional brand of a newspaper.
Unlike social media, AI chatbots have yet to face major content moderation crises. This is attributed to a temporary grace period where users understand the technology is flawed and prone to 'hallucinations.' This widespread forgiveness lowers the stakes for errors, but expectations will rise as AI matures.
Unlike social media's ad model, AI's revenue is driven by enterprise clients who demand accuracy for business use. This creates a powerful financial incentive for AI labs to prioritize truthfulness over virality, a complete reversal from the social media era that could benefit information quality.
The AI industry has no third-party verification; labs self-report performance on bias and accuracy via blog posts. Campbell Brown likens this to banks auditing themselves, arguing it creates an accountability vacuum and undermines public trust in a foundational technology.