A business’ guide to preventing online fraud

Xavier Sheikrojan at Signifyd explains how fraud tactics are evolving through the use of AI

Businesses are at a crossroads. Regardless of industry, companies are having to contend with the exponential innovation brought on by emerging technologies and connectivity online. 

While this connectivity undoubtedly represents a great opportunity for businesses, it also presents challenges and risks. These include the evolving fraud landscape, as fraud becomes less opportunistic and more organised in nature. 

The way we’re seeing fraud evolve most worryingly includes the implementation of smart technologies like AI to create chatbots and deepfakes that trick businesses into giving up sensitive information, ultimately exposing holes in security frameworks and opening up huge financial and reputational risk. 

AI fraud represents a significant threat to businesses’ bottom line, so it is vital to not only know what scams fraudsters are using, but also how companies can fight back.

An evolution in cyber-crime

In the early days of the internet, cyber-crime (and fraud) was relatively tame; fraudsters might impersonate a legitimate organization such as His Majesty’s Revenue and Customs (HMRC) and pretend tax money was owed, allowing them access to card details. While this sounds scary, there were typically tells for this sort of scam – the email from “HMRC” would contain spelling and grammatical errors, and would read that urgent action needs to be taken today.

That soon changed though, and online fraud is now a different beast altogether. As the general public became more aware of these scams, crime evolved. While we once saw individual fraudsters trying to make a quick buck off of the innocent and ignorant, now organised teams dedicate themselves to learning, pivoting, and adapting their strategies. 

An example we’ve seen of crime syndicates that specialise in online fraud is the Southeast Asian fraud ring, which in 2022 stole a staggering $660mn USD in phones, laptops and microchips. This event, and others like it, illustrates the need for continual innovation in fraud prevention.

Today, the use of AI is endemic among cyber-criminals. Not only are these technologies cheap to use, they’re also difficult to trace or pre-empt. 

Using AI to unmask cyber-crime

We’re seeing AI used broadly across cyber-crime, but within the fraud space, chatbots and, more worryingly, deepfakes pose a real threat to businesses. By taking in massive amounts of data—texts, emails, etc.—and recreating genuine-seeming conversations, chatbots are avoiding the classic tells of fraudsters in the past, learning to replicate real communication styles that are difficult to detect as fake. 

While chatbots recreate text, deepfakes and synthetic personalities are taking impersonation to a whole new level. Used to generate digital likenesses and voices of real people, the potential of this technology to generate entire fake personas is a huge risk to businesses.

As AI continues to evolve and improve, it will only get harder for businesses to spot criminals for what they are. The only way to fight back effectively is to use cyber-criminals’ own tools against them, and fight AI with AI.

Crimestoppers: backing AI with experts

We’ve all seen AI’s ability to learn and adapt as inputs increase. Harnessing this feature means businesses tapping into data reserves; feeding behavioural and identified crime data into their own models to create a feedback loop for their friendly AI. The more information that is input, the more AI is able to observe and identify crime while it occurs. 

AI is a huge leap forward in technology solutions, but it isn’t a panacea. AI is still a student and all students need teachers. 

Enter intelligence teams and fraud prevention network partners.

The role of an intelligence team is to evaluate existing crime cases on behalf of businesses, collaborating closely with them on data science analysis. These groups are responsible for feeding information into businesses’ AI models to encourage learning and improvement.

They can also make use of their own AI models in cases where businesses cannot create or invest in their own machine learning or AI solutions.

Intelligence teams are also responsible for overseeing the use of data from fraud prevention networks. These networks allow businesses to tap into huge volumes of data from real people, allowing for better vetting of authentic vs. fraudulent interactions over text or online.

Working with a fraud prevention network often means they will take on the risks associated with fraudulent AI. As online fraud continues to evolve, this is an increasingly valuable tool to integrate into businesses’ back ends.

To ensure that businesses are keeping ahead of the crime curve, implementing these solutions and systems is a critical piece of the puzzle to not only prevent attacks, but also de-risk losses should crime occur.

AI is the key to fighting cyber-crime, but without a strong network and data analysis, it’s like having car keys but being unable to start the engine. By working with the right people, and making use of friendly AI, businesses can free people up to act with confidence, whatever their business. 

Xavier Sheikrojan is senior risk intelligence manager at Signifyd

Main image courtesy of iStockPhoto.com and Love portrait and love the world