To Misinformation Researchers, AI Is a Scourge—And a Powerful New Tool

Why It Matters

AI is amplifying the scale and sophistication of disinformation, threatening democratic discourse, while simultaneously offering researchers unprecedented analytical capabilities to combat it.

Key Takeaways

• AI boosted disinformation output without reducing credibility, per DCWeekly case
• EU report: AI use in foreign influence hit 27% in 2025
• Researchers employ LLMs to detect and rerank polarizing social‑media content
• AI‑driven bot swarms threaten democratic discourse by mimicking authentic conversations
• Funding shifts favor AI projects as disinformation grants face U.S. cuts

Pulse Analysis

The rise of generative artificial intelligence has turned the misinformation landscape into a high‑speed production line. A 2025 European Union analysis found that AI was behind 27% of foreign influence operations, a three‑fold jump from the previous year, underscoring how quickly bad actors can scale tailored narratives. This surge is not limited to state actors; the Russian disinformation outlet DCWeekly.org switched to AI‑generated articles, doubling its output while preserving the same level of perceived credibility, illustrating the technology’s efficiency as a content‑creation engine.

At the same time, the same AI tools are reshaping how scholars study false information. Large‑language models enable researchers to sift through millions of posts, flag subtle misinformation, and even manipulate experimental feeds to reduce polarization. Experiments using AI‑driven browser extensions have shown that lowering exposure to polarizing content can improve attitudes toward opposing political groups, a result that would have been impossible without automated text analysis and real‑time content reranking.

Policy implications are profound. As AI‑generated bots infiltrate platforms like X, they can masquerade as genuine users, creating the illusion of consensus and overwhelming fact‑checkers. Funding trends reflect this tension: U.S. agencies are cutting traditional disinformation grants while still supporting AI research, prompting scholars to pivot toward AI‑focused proposals. The field must therefore develop new guardrails—both technical, such as watermarking AI output, and institutional, like transparent data‑sharing agreements—to keep pace with an ecosystem where the tools of deception and detection are increasingly the same.