Climate Misinformation Expected to Surge on Social Media, Expert Warns

Meta’s Fact-Checking Rollback Raises Alarm Over Climate Misinformation

Meta’s recent decision to end its fact-checking program in the United States has sparked concerns that climate misinformation could proliferate unchecked across Facebook and Instagram platforms. The tech giant will terminate agreements with U.S.-based third-party fact-checking organizations in March 2025, potentially creating an information vacuum during climate disasters when accurate information is most critical.

Since 2020, Meta has utilized its Climate Science Information Center to combat climate misinformation on Facebook. The company’s current approach involves third-party fact-checkers flagging false and misleading posts, after which Meta decides whether to attach warning labels or reduce algorithmic promotion of such content. The policy prioritizes identifying “viral false information,” hoaxes, and “provably false claims” while excluding opinion content.

The announced changes will only affect U.S. users, as Meta will continue fact-checking content viewed internationally. This geographical distinction stems from stronger regulations combating misinformation in other regions, particularly the European Union.

Climate communication researchers have found that fact-checks play a crucial role in correcting misinformation, though their effectiveness varies based on people’s existing beliefs, ideology, and knowledge. Finding messages that align with audience values and using trusted messengers significantly improves outcomes.

The timing of this policy shift is particularly concerning as extreme weather events linked to climate change become more frequent and severe. Heat waves, flooding, and wildfires are increasingly common and devastating as global temperatures rise. Research shows these disasters typically trigger spikes in social media attention to climate change, creating prime conditions for misinformation to spread.

The challenge has been compounded by the rise of AI-generated content. During recent disasters like Hurricanes Helene and Milton, fake AI-generated images went viral, hampering the Federal Emergency Management Agency’s disaster response efforts. These incidents highlight the difference between misinformation—false content shared without intent to mislead—and disinformation, which is deliberately deceptive.

Organized disinformation campaigns are already occurring. Following the 2023 Hawaii wildfires, researchers from multiple organizations documented a coordinated propaganda campaign by Chinese operatives targeting U.S. social media users.

Meta CEO Mark Zuckerberg has specifically cited X’s Community Notes as inspiration for his company’s planned content moderation changes. However, research suggests crowd-sourced fact-checking approaches are too slow to effectively counter viral misinformation during its peak spread. By the time community corrections appear, false claims may have already reached millions of users.

Climate misinformation poses particular challenges because it tends to be “sticky”—difficult to dislodge once people encounter it repeatedly. Research indicates that simply sharing more facts is ineffective against false climate claims. Instead, explaining the scientific consensus on climate change in advance can “inoculate” people against misinformation.

The U.S. public generally supports industry efforts to moderate false information online. Instead, tech companies are increasingly shifting responsibility to users. During climate disasters, when people need reliable information to make potentially life-saving decisions, crowd-sourced debunking may prove inadequate against organized disinformation campaigns.

As seen recently when Los Angeles County erroneously sent an evacuation alert to 10 million people in January 2025, communication during disasters is already challenging. Without professional fact-checking systems in place, conditions for the rapid spread of false content could significantly worsen under Meta’s new approach to content moderation.