Study Warns of AI-Fueled Fake News Threatening Bank Runs

​A new study by UK research company Say No to Disinfo and communications firm Fenimore Harper highlights the growing risk of bank runs triggered by AI-generated fake news disseminated on social media. The study suggests that generative AI can be used to create convincing false narratives about a bank's financial stability, potentially leading to rapid withdrawals by depositors.

The ease and speed with which AI can generate and spread disinformation are key concerns.  Researchers showed sample AI-generated content, such as fake news stories claiming customer money is not safe or memes joking about security issues, to UK bank customers. They found that a third were "extremely likely" to move their money after seeing it, with a further 27% "somewhat likely".

"As AI is making disinformation campaigns easier, cheaper, quicker and more effective than ever before, the emerging risk to the financial sector is rapidly growing but often overlooked," the report stated. The study emphasizes the potential for malicious actors to exploit AI to create and amplify damaging narratives about financial institutions.

The study estimated that for every £10 spent on social media advertising to amplify fake content, as much as £1 million in customer deposits could be at risk. This calculation, based on average UK customer deposits, social media ad costs, and estimated viewership, illustrates the potential scale of the problem.

The report recommends that banks implement enhanced monitoring of media and social media, integrated with withdrawal tracking systems, to detect the impact of disinformation on customer behavior.  This integrated approach would allow banks to identify unusual withdrawal patterns that correlate with the spread of negative information online.

While some institutions, like Revolut, have acknowledged the potential threat and emphasized their existing monitoring practices, others, including NatWest and Barclays, declined to comment.  Revolut's head of financial crime, Woody Malouf, said the fintech conducts real-time monitoring for emerging threats and believes that while an industry-wide event is unlikely, it's still possible and requires preparedness. He also stressed the need for social media platforms to take greater responsibility in preventing the spread of disinformation.

Industry body UK Finance noted that "banks are working hard to manage and mitigate risks around AI and the regulatory authorities are looking at the potential financial stability challenges the technology poses." This suggests that while banks are aware of the risks, the specific challenge of AI-generated disinformation requires further attention and coordinated action.  The report's release coincided with an AI Summit in France, where the focus was on promoting AI's spread, a shift from previous summits that emphasized risk management, highlighting the evolving conversation around AI's impact on society.