Five Key Insights From Full Fact’s 2025 Fact-Checking Efforts

Full Fact Reflects on 2025: Key Trends in Misinformation and Political Claims

As the team at Full Fact wraps up for their Christmas break, they’ve been reflecting on a busy year of fact-checking that saw them publish over 750 fact checks and related content. From major political disputes to alarming online falsehoods, their work has covered a broad spectrum of misinformation circulating throughout 2025.

The organization’s review highlights five key trends that shaped the misinformation landscape this year and offers insights into what might lie ahead in 2026.

AI has dramatically accelerated the spread of misinformation, according to Full Fact’s analysis. What was once a novelty in their fact-checking work has become routine. In November 2024, the organization suspected AI involvement in just four published fact checks. By October 2025, that number jumped to at least 27, with AI frequently used to generate false claims about government policies and personal freedoms.

“Generative AI has gone from being a novelty in our fact checking to absolutely routine,” the report states, suggesting this trend will intensify as AI tools become more convincing and widely adopted.

The political sphere has seen a notable shift, with Reform UK emerging as a key player in fact-checking battles. With the party rising in the polls, claims from Reform UK about immigration issues have become some of Full Fact’s most-read content. Simultaneously, fact checks about claims made against Reform UK politicians have gained significant attention, with the organization’s most popular fact check of 2025 examining whether Labour was correct in saying Nigel Farage’s party would “scrap the NHS.”

As Scotland, Wales, and parts of England prepare for May 2026 elections, Full Fact anticipates that claims and counterclaims between Labour and Reform UK will intensify significantly.

Another interesting development is the growing reliance on AI tools for fact-checking itself. Social media users frequently call on AI assistants like Grok to verify information, creating a new dimension in the fight against misinformation. However, Full Fact’s investigations have uncovered numerous instances where these AI tools provided inaccurate responses, including Google Lens sharing misleading information about images and Grok wrongly attributing the timing of a police video.

As the use of AI for verification continues to grow, questions remain about whether these tools will become more reliable and whether users will develop better skills for evaluating their responses.

Full Fact has also expanded its Government Tracker, which now monitors 86 Labour pledges. This work has highlighted the importance of scrutinizing the fine print of political commitments. Their investigations revealed that government celebrations of increased hospital appointments masked the fact that the growth was actually smaller than under the previous Conservative administration. Similarly, their detailed analysis of Labour’s tax pledges allowed for clear assessment of whether the November Budget breached manifesto commitments.

Finally, the organization continues to battle “zombie claims” – persistent falsehoods that resurface repeatedly despite being debunked. These range from old myths about MPs’ breakfast expenses to newer misconceptions about unemployment figures under Labour. While Full Fact has achieved some successes in combating these persistent falsehoods – including securing corrections from the House of Commons and Lords libraries, the Scottish government, and major medical journals – they acknowledge that eliminating entrenched misinformation remains an ongoing challenge.

As they look toward 2026, Full Fact’s team remains committed to tackling these persistent claims, recognizing that doing so is crucial not only for public information but also for improving the accuracy of large language models that increasingly shape our information ecosystem.