The Police and Crime Committee convened to discuss the Metropolitan Police Service's (MPS) efforts in combating online fraud in London. Key concerns raised included the increasing sophistication of AI-enabled scams, the significant under-reporting of fraud, and the emotional and psychological toll on victims. The committee also explored the challenges in police recruitment and retention for specialist cybercrime roles, and the effectiveness of current fraud prevention campaigns.

Online Fraud in London: Scale, Typologies, and Emerging Threats

Detective Superintendent Oliver Little of the City of London Police (CoLP) highlighted that London accounts for approximately 40% of national fraud victimisation. He noted the transition from Action Fraud to the new Report Fraud system, designed to centralise reporting and improve the victim journey. Will Lyne, Head of Economic and Cybercrime for the MPS, added that nationally, fraud reports saw a 1% increase in 2024-25, while cyber-dependent crimes rose by 60%. He estimated that around 55,000 fraud and 10,000 cybercrime reports originate from London annually.

Ruth Halkon, Research Manager for The Police Foundation, emphasised that fraud exists on a continuum, with up to 70-80% of frauds having an online element. Rory Innes, CEO of The Cyber Helpline, stated that only about 36% of victims report to the police, suggesting that reported figures are likely a third of the actual number of victims. He stressed that the impact of fraud extends beyond financial loss to mental health and online confidence.

The panel identified AI-enabled scams and deepfake impersonation as significant emerging threats. Detective Superintendent Little explained how technology allows fraudsters to attack a wider audience and layer deception, citing Hi Mum scams as an example where voice cloning could make manipulation easier. Will Lyne detailed six priority fraud types in London: romance fraud, courier fraud, investment fraud, abuse of position fraud, payment diversion, and card and payment fraud. He also highlighted three key trends: the hybridisation of fraud types, the increasing use of cryptocurrency, and the role of AI in optimising and scaling fraud operations.

Rory Innes described criminals as being better at marketing than they are at technology, understanding what makes individuals believe them and click. He warned of the potential for AI to enable chatbots to deliver scams at the volume previously requiring call centres, lowering the barrier to entry for offenders.

Perpetrators and Victims of Online Fraud

The discussion revealed that victims of online fraud are diverse, spanning all demographics. While stereotypes of elderly individuals are common, research suggests those in their 40s to 60s are highly vulnerable. There is also a rising number of young people being targeted, with 88% of those under 21 surveyed by CREST and CoLP having encountered fraudulent content online, yet only 6% reporting it. Young people often perceive these scams as normal, leading to under-reporting and a lack of help-seeking.

Will Lyne noted that while older individuals are more likely to report, online fraud affects all age groups, with a significant problem among 13-20-year-olds, particularly males, concerning sextortion. He shared a tragic case where a mother lost her son to suicide over a £200 sextortion scam. Businesses, especially SMEs, are also vulnerable due to limited resources for protection measures.

Legislation, Reporting, and Victim Support

Ruth Halkon indicated that legislation is struggling to keep pace with technological advancements, with the Online Harms Bill making some reference to deepfakes but generally playing catch-up. She advocated for an international framework to tackle issues like deepfakes. The panel agreed that a more victim-centric approach is needed in reporting systems, with improved questioning to avoid deterring victims.

The significant issue of under-reporting was repeatedly highlighted. Rory Innes stated that only about a third of fraud cases are reported, and of those reported to Action Fraud/Report Fraud, 87% receive an NFA decision, rising to 98.1%. This lack of perceived action discourages reporting. He stressed the need for a middle step to help victims and improve the quality of reports to enable investigations.

The emotional and psychological impact of fraud was a major theme. Ruth Halkon's research found that 92% of fraud victims reported mental, physical, or behavioural symptoms, with stress and anxiety being common, but also physical health symptoms and significant impacts on daily life. She described fraud as a hidden public health crisis. Rory Innes echoed this, noting that even when financial loss is not substantial, the impact on mental health and online confidence is profound.

The role of The Cyber Helpline was discussed as a vital bridge, providing expert support and guidance, particularly for those who are hesitant to report to the police due to shame, fear of retribution, or lack of technical understanding. The potential impact of increased funding for The Cyber Helpline was highlighted as a cost-effective way to scale support and improve reporting quality.

Police Response and Collaboration

The MPS and CoLP are working closely together, with CoLP leading nationally on economic crime. They are utilising AI and automation in their own systems to detect and respond to fraud, and collaborating with the private sector, including banks, through initiatives like the banking protocol and the DCPCU. The MPS is also focusing on proactive operations, such as Operation Callback against courier fraud, which reportedly halved offending and reduced victim losses by two-thirds.

Challenges in recruiting and retaining specialist cybercrime officers were acknowledged, with the private sector offering more lucrative salaries. However, the MPS and CoLP are exploring innovative training pathways and programmes like Police Now to attract talent. The MPS is also investing in its economic crime teams, with an uplift in staff funded by the Asset Recovery and Centralisation Scheme (ARIS).

The effectiveness of fraud prevention campaigns is being evaluated through impact-focused metrics, aiming to understand behavioural changes rather than just reach. The Be Switched On campaign was cited as an example of targeted messaging to encourage security feature activation on phones.

The Role of AI and Future Challenges

The panel expressed concern about AI's potential to increase the volume and sophistication of online fraud. While AI is not yet running end-to-end fraud models, it is being used to optimise and enhance various stages of fraud operations. The ease with which AI tools can be rented lowers the barrier to entry for criminals. The development of defensive AI tools and the establishment of the Online Civic Culture Centre (O3C) concept, a public-private partnership for law enforcement, were discussed as ways to combat these evolving threats.

The discussion also touched upon the need for better legislation, potentially an international framework for deepfakes, and a more proactive approach to technology release, akin to safety testing for cars or pharmaceuticals. The challenge of identifying and recording AI involvement in fraud cases was noted, with victims often unaware of the technology used against them.

Collaboration and Future Directions

The importance of collaboration between law enforcement, the private sector, and charities was repeatedly stressed. The new Report Fraud system aims to be victim-centric, with improved questioning and embedded victim services. The panel also discussed the need for consistent messaging and trusted sources of advice, potentially leveraging influencers and community leaders to reach vulnerable demographics. The potential for targeting educational messages at critical life moments, such as opening a bank account or applying for a mortgage, was also raised.

The meeting concluded with a note of the committee's work programme and the date of the next meeting. The discussions underscored the multifaceted nature of online fraud and the ongoing need for a coordinated, victim-focused, and technologically adept response.