How Meta’s fact-checking changes could fuel climate misinformation on Facebook and Instagram

Meta, the parent company of Facebook and Instagram, has decided to wind down its fact-checking programme and scale back content moderation more broadly. That move raises an obvious question: what will people actually see on those platforms from here on?

A particular concern is that the shift could allow far more climate misinformation to circulate across Meta’s apps, including claims that are misleading or ripped from context during disasters.

Back in 2020, Meta launched its Climate Science Information Center on Facebook as a response to climate misinformation. At present, third-party fact-checkers partnered with Meta identify posts that are false or misleading. Meta then determines whether to add a warning label and to limit how heavily its algorithms amplify those posts.

Under Meta’s current approach, fact-checkers are directed to prioritise "viral false information," hoaxes and "provably false claims that are timely, trending and consequential." Meta also makes clear that this does not cover opinion content, so long as it does not contain false claims.

Meta will terminate its agreements with US-based third-party fact-checking organisations in March 2025. These planned changes, which are due to roll out to US users, will not affect fact-checking content seen by users outside the US. In many other places, including the European Union, the tech industry faces tighter regulation aimed at tackling misinformation.

Fact-checking curbs climate misinformation

I research climate change communication. Fact-checks can correct political misinformation, including misinformation about climate change. How effective they are depends in part on people’s beliefs, ideology and what they already know.

It can help to frame messages so they fit the values of the audience, and to use trusted messengers - for instance, climate-friendly conservative groups when speaking to political conservatives. Appealing to shared social norms can also be effective, such as the idea of reducing harm to future generations.

As the planet warms, heatwaves, flooding and conditions that fuel fires are becoming more frequent and more devastating. When extreme weather strikes, social media attention to climate change often surges. Posting tends to spike during a crisis and then tail off rapidly.

Confusion online during emergencies is being worsened by low-quality fake images produced with generative artificial intelligence software - so-called AI slop. For instance, after back-to-back hurricanes Helene and Milton last autumn, an AI-generated image of a young girl, shivering and clutching a puppy in a boat, spread widely on the social media platform X. Meanwhile, the circulation of rumours and misinformation impeded the Federal Emergency Management Agency’s disaster response.

The key difference between misinformation and disinformation is the intention of whoever is sharing it. Misinformation refers to false or misleading content shared without a deliberate aim to mislead. Disinformation, by contrast, is false or misleading information shared specifically to deceive.

Disinformation efforts are already under way. Following the 2023 Hawaii wildfires, researchers at Recorded Future, Microsoft, NewsGuard and the University of Maryland separately documented a co-ordinated propaganda campaign by Chinese operatives aimed at US social media users.

Of course, the spread of rumours and misleading information on social media is not a new issue.

Yet moderation strategies do not all have the same impact, and platforms are changing how they handle misinformation. X, for example, replaced its rumour controls - which had helped to debunk false claims during rapidly unfolding disasters - with Community Notes, a system of user-generated labels.

False claims can go viral rapidly

Meta CEO Mark Zuckerberg has pointed to X’s Community Notes as a model for his company’s planned moderation changes. The problem is that false claims can become viral very quickly.

Recent research shows that crowd-sourced Community Notes often appear too slowly to halt the spread of viral misinformation early in its online life cycle - the stage when posts receive their widest exposure.

With climate change specifically, misinformation is "sticky." Once people encounter the same falsehood repeatedly, it becomes particularly difficult to remove from their thinking.

Climate misinformation also erodes public acceptance of established science. Simply providing additional facts is not enough to stop false climate claims from spreading.

One way to help people resist misinformation is to explain that scientists agree climate change is occurring and is driven by humans burning greenhouse gases. Psychology research suggests this "inoculation" approach reduces the influence of claims that say otherwise.

That is why warning people about climate misinformation before it becomes viral is so important for limiting its reach. On Meta’s apps, that task is likely to become more difficult.

Social media users as sole debunkers

With these forthcoming changes, the responsibility for fact-checking on Facebook and other Meta apps will fall to you. The most effective way to pre-bunk climate misinformation is to begin with accurate information, then briefly flag the myth - and mention it only once. After that, explain why it is wrong and repeat the truth.

During climate change-fuelled disasters, people urgently need trustworthy information to make decisions that can save lives. That is already hard enough, as shown when the Los Angeles County emergency management office mistakenly sent an evacuation alert to 10 million people on 9 January 2025.

Crowd-sourced debunking cannot match organised disinformation campaigns when a crisis creates an information vacuum. With changes to Meta’s moderation policies and algorithms, the conditions that allow misleading - and outright false - content to spread rapidly and unchecked could become even worse.

Overall, the US public wants the industry to moderate false information online. Instead, it appears that major tech companies are shifting the burden of fact-checking on to their users.

Jill Hopke, Associate Professor of Journalism, DePaul University

This article is republished from The Conversation under a Creative Commons licence. Read the original article.