Fraud Detection in a World of Deepfakes

Firms need to be diligent to avoid being the victim of fraud, particularly when receiving fund transfer instructions from clients via email. The prudent advice provided by law societies and/or lawyer’s insurance when transferring funds is to verify emailed instructions through direct telephone or in-person contact. The telephone contact should be initiated by the firm by using the number on file to avoid the communications being intercepted by a fraudster.

Two recent examples highlight this importance. The Lawyers Indemnity Fund in B.C. recently reported that a firm’s diligent adherence to protocol saved them from a funds transfer fraud. When a paralegal at the firm received fund transfer instructions from a client via email, the paralegal brought the request to the attention of a lawyer at the firm as per the firm’s protocol. The lawyer telephoned the client to confirm the transfer and instead discovered the email was fraudulent. A firm in the prairies faced a similar circumstance when acting for a corporate client selling property. However, the assistant who received instructions via email did not telephone the client to confirm. Over $250,000 was fraudulently transferred.

The rise of AI may provide numerous opportunities for firms to streamline processes and deliver more cost-effective services to their clients. It may also provide more opportunities for fraudsters. One potential area of concern is the use of deepfakes to dupe unsuspecting firms. A deepfake uses AI to generate or manipulate an image, audio or video content to appear realistic when it is not authentic.

As noted in a blog post entitled AI-Deep Synthesis Regulations and Legal Challenges: Recent Face Swap Fraud Cases in China, the legal representative of a Chinese technology company was recently fooled into transferring 4.3 million yuan by AI-powered face swapping technology during a video chat with a client. Voice emulators that can replicate a person’s voice with a voice sample are also freely available on the internet.

Moving forward, how will we trust the instructions we receive from clients if fraudsters increasingly utilize deepfake technology? Continued reliance on low-tech options like receiving in-person confirmation might be the safest bet but AI tools to detect fraud also appear to be on the rise. Stay tuned for more developments and guidance as we move to stay a step ahead of fraudsters.

The post Fraud Detection in a World of Deepfakes appeared first on Slaw.

Related Posts