AI had such promise, yet right now it is very much being used for bad things. A look at the news this week reveals that fake traffic, fake data, fake personas and fake dating profiles proliferate – and cost everyone except for the scammers dearly. And it is all thanks to AI.
For example, a recent report from TrafficGuard concludes that top tier sports betting sites see a 44% invalid traffic rate from Google, wasting some 40% of their ad spend.
Meanwhile, MCP Insight has launched DATESCAN to help the dating industry understand flows, efficiency and more, which will certainly help with understanding how these sites are being used by scammers.
We also learn how businesses need to work harder on complying with the latest raft of PCI DSS regs – which also feature guidance on how to combat AI powered phishing fraud – while the University of Portsmouth is offering a raft of new tools to help combat AI-powered deepfakes.
If it wasn’t for AI being deployed to make CPaaS more effective, it would be easy to think that AI was the enemy…
AI had such promise, yet right now it is very much being used for bad things. A look at the news this week reveals that fake traffic, fake data, fake personas and fake dating profiles proliferate – and cost everyone except for the scammers dearly. And it is all thanks to AI.
For example, a recent report from TrafficGuard concludes that top tier sports betting sites see a 44% invalid traffic rate from Google, wasting some 40% of their ad spend. Tier 2 operators, regional heavyweights with leaner operations, experience a 29% IVT rate, with 17% of their ad budgets dissipated by invalid traffic, while tier 3 contenders, mid-sized players in the market, face a 42% IVT rate, sacrificing 37% of their advertising funds to these inefficiencies. Tier 4 operators, the smaller underdogs, contend with a 33% IVT rate, losing 19% of their ad spend to similar issues.
The research also finds that 76.8% of total website traffic for sports betting operators originates from returning users. This high proportion indicates that navigational traffic is infiltrating paid campaigns, diverting resources intended for new user acquisition and potentially hindering market growth.
And it is all down to AI.
Meanwhile, deep fake images and personas are a big headache for all businesses as they allow for fraud to be conducted at scale and to appear just like real people are doing the interacting. From identity theft right through to impacting how people get scammed on dating sites, there is an almost out of control trend for creating ‘real’ people that are very much not human and which are then used to scam millions from businesses and ordinary people alike.
And it is all down to AI.
In fact, if there is a scam out there – old or new – then it has been either created, given a new lease of life or greatly enhanced by artificial intelligence. It is a technology that is becoming a runaway success and doing bad things.
AI solutions abound
Of course, the wider tech industry is fighting fire with fire, using AI to try and beat these scams and most of the above stories we have covered this past week are actually from companies that have tools to help make these things work better.
TrafficGuard, for example, is doing much to help sports betting sites – and many others – filter out this ‘bad traffic’ in an attempt to make the multi-billion dollar sports betting market work better.
Likewise, the UK’s University of Portsmouth’s Artificial Intelligence and Data Science (PAIDS) Research Centrehas created DeepGuard, AI based software that has been trained on deepfakes to look out for the (very) subtle glitches and inaccuracies in deepfake images and videos to try and weedle out those that are problematic.
On the dating front, MCP Insight and the Online Dating and Discovery Association to create DATESCAN for the dating industry. While not specifically aimed at discovering fraud, it is a tool that is used to test dating services for efficiency, flow and compliance across many markets, but that in itself touches on fraud and other scams around the dating business.
Of course, across the telemedia industry there are the likes of Evina, Tetra Mobile and Empello testing and measuring and checking the compliance of myriad services, all helping to uncover and block fraud. In fact, the telemedia sector – once a hot bed of naughty services that sailed very close to the wind of legality – is perhaps now one of the most on the ball when it comes to using AI to fight this wealth of AI scams.
AI out of control?
But I can’t help feeling that AI-powered fraud is running out of control. While the rapid rise of AI in 2023 and 2024 prompted much handwringing over the machines taking over, designing evermore powerful AI that would soon outstrip human intelligence – and that is still very much a threat – in the short term, it seems that AI fraud is a much bigger problem.
While the machines are busily creating more powerful AI that will eventually be so clever us mere meatbags won’t be able to understand it, the human fraudsters are leveraging off-the-shelf and readily available, open-source tech to create fake people, fake IDs, fake traffic and a raft of scams that will cost us all so much money we may yet welcome the rise of the uber-machines. Perhaps only the AI-son-of-AI will be able to finally stop the fraudsters?
