Revealing AI: the Dual Potential of Artificial Intelligence in Combating - and Enabling - Fraud

Advancements propelled by AI are occurring rapidly. Not only are tangible outcomes consistently revealed, but artificial intelligence has become mainstream, dominating headlines and finding increasing utility in everyday life.

It is undeniably an exciting era for professionals in healthcare, data science, and numerous other fields bolstered by AI.

However, swift progress brings its own set of challenges. The upcoming EU AI Act is rumored to be challenging to define, given the rapidly changing landscape of possibilities. The pervasive issue of misuse also looms large. Despite its myriad benefits, new technology is often exploited for malicious purposes, often due to the absence of regulations during its inception.

Consider the early days of the internet as an example. Even before the World Wide Web's debut, and when the internet was primarily used by academic researchers, the first major cyber attack was executed by Cornell grad student Robert Morris in 1988. Dubbed the 'Morris Worm,' it marked the first felony conviction in the US under the 1986 Computer Fraud and Abuse Act, sparking a technological arms race between hackers and developers for years to come.

AI now finds itself at a similar juncture. With the rise of deepfake technology and increasingly sophisticated scams, those championing ethical AI must navigate how to minimize fraudulent activities and, ultimately, harness AI for benevolent purposes.

Escalating Sophistication

AI scams are proliferating, and so is the number of individuals falling victim to them. Criminals exploit the ability to replicate video or voice notes, resembling loved ones, resulting in one in four surveyed individuals having fallen victim to an AI scam, according to a McAfee study. Notably, 70% of respondents lacked confidence in distinguishing a fake AI-generated voice from a real one.

As AI evolves, scams take diverse forms. A high-profile case involved financial journalist Martin Lewis in 2018. Exploiting his trusted status due to the Money Saving Supermarket website, criminals created a deepfake video on Facebook, where Lewis seemingly encouraged consumers to part with their money. With robust regulation not yet in place, legal responsibility remains unclear, emphasizing the need for prompt legislative action.

Unexpected Benefits

Despite AI's misuse, its exploitation by criminals has some unexpected benefits for developers. Similar to the 'Morris Worm,' new threats force innovators to act swiftly, developing rapid solutions and advancing AI to counter dangers. This adds an additional layer of protection for the public.

While AI growth leads to misuse, it also fosters expertise as developers and criminals vie for supremacy. The technological cycle, driven by the unprecedented openness of AI techniques and open-source software, encourages collaboration, crucial for addressing challenges like fraud.

Fraud Prevention

Ironically, despite misuse, AI is increasingly effective at preventing fraud. Particularly in banking, Juniper Research predicts that global spending on AI-enabled financial fraud detection will exceed $10 billion by 2027. Banking institutions are already implementing AI-powered preventive solutions, such as Mastercard's Consumer Fraud Risk, adopted by nine UK banks. AI enables swift detection and prevention of fraudulent activities, saving substantial sums.

While AI faces ambiguity and exploitation, it paradoxically offers solutions for fraud prevention. The upcoming EU AI Act may address challenges, but until then, the ethical AI movement must find innovative solutions, prioritizing collaboration and openness. With preventive measures and enhanced public education, developers can mitigate the impact of fraudsters on unsuspecting victims.