Platforms realize they don’t need journalists

Tech companies will double down on improving the quality of information they provide in 2026, but that investment won’t help publishers find a path to stability.

The good news is that tech companies will face pressure in the year ahead to bolster the information ecosystem. As competition heats up between OpenAI, Google, Anthropic and other tech companies, there will be real pressure to create chatbots and agents that do more than write bad poetry. Those giants know their success is dependent on convincing people their products are trustworthy and reliable.

That means we’ll see greater emphasis on building products that deliver real utility. With nearly a billion people around the world using AI every week, the cost of high-visibility mistakes — like telling people to eat rocks or encouraging your customers to kill themselves — will drive the platforms to invest in trust and safety initiatives that protect their bottom line.

The bad news, for news organizations at least, is that tech companies will realize they don’t need journalism to give people the answers they need.

To be sure, tech companies need vast quantities of ground truth information — verified data that can be used for training or enhancing generated responses with techniques like RAG. A handful of global publishers will benefit from ongoing licensing deals that let AI platforms leverage their reporting to summarize current events in answer to direct news queries.

But companies like OpenAI and Google simply don’t need that much journalism to meet their users’ needs: Searching for news will remain a small subset of the broader applications of generative AI. A research paper from OpenAI showed that about 19% of interactions with ChatGPT from May 2024 to June 2025 were seeking specific information; asking about the news is likely a fraction of that.

Tech companies are more likely to seek ground truth through paths that don’t depend on newsrooms. Rather than ingesting your health care reporting, they’ll go straight to the World Health Organization to source raw data from researchers around the world. Instead of waiting for a breakthrough to get reported, they’ll scrape Arxiv and synthesize their own coverage.

On some beats, journalists and tech companies will compete for access to the same data sources: Nothing stops Google from directly quoting legislative activities posted to congress.gov instead of waiting for a bill’s introduction to be reported online.

In other domains, the outcome could be far more dystopic: If a tech company licensed CCTV footage, it could plausibly use that to answer basic questions about things happening in a community. Without editorial standards, however, there’s no telling how that footage might be used.

This path is fraught for many reasons, not least of which is that it underscores the threat facing newsrooms: It’s not just that fewer people will come directly to news websites, but that generative AI poses a direct challenge to the role we serve for our audience.

There may be an upshot if the competitive pressures on AI platforms protect the quality of information online — an essential step toward rebuilding trust in civil society. That would be good news for democracy, even as it doesn’t answer the question of how newsrooms will become sustainable.

So what can we do?

We can double down on talking to our audience to understand their needs. We can recommit our organizations to serving clear communities. We can relentlessly scrutinize our work for procedures and content types that don’t make sense anymore.

The threats we face are existential, but we can reframe them as opportunities.