PartyGaming has adopted and refined a new method for affiliate monitoring, after introducing experimental systems using artificial intelligence.
The system, created by Envisional, aims to toe the line between managing risk from affiliates, preventing security lapses, and encouraging entrepreneurialism, according to Ian Shircore, marketing director at Envisional.
Shircore: “Maintaining security with affiliates is not just an anti-fraud measure, there is a wider issue. That is: no country will think to regulate online gaming while companies cannot keep track of their affiliates.”
The issue of violations committed through affiliates is one which was highlighted at last week’s Combating Cybercrime conference. Examples of money laundering through affiliates were given by speakers including representatives from PartyGaming and Unibet, as well as the UK’s Serious Fraud Office. One attendee said that the industry had been “dragging its feet” in terms of affiliates, and that greater security was needed if the sector was to show its maturity.
However, it was also noted that the importance of affiliates for operators should not be underestimated, and that minor “offences” from what were termed “over-enthusiastic” affiliates, are dealt with most effectively by operators.
Shircore said: “If you think of how affiliate networks work, they are the very essence of entrepreneurial spirit. Nobody wants to crush that, and it will save time for the companies to just have to recognise what’s going on, and say, please don’t step over the line.
“In terms of affiliate monitoring, people desperately need to know what’s going on. The enforcement is pretty simple. The relationship between the operator and the affiliate is not necessarily damaged.”
Shircore added that the collaboration with Party had resulted in a product that was unusual for its use of artificial intelligence.
He said: “We are pushing the limits of artificial intelligence, and we have now a system with greater flexibility for the operator. The artificial intelligence follows links like a human, and uses human-like logic. It also searches images. We produced prototype systems which we worked on for PartyGaming. It’s been remodelled and refined slightly ‘in reality’ but it is basically the same concept.
“What they get is a pyramid of results – and the pyramid will show violations which at the top of the pyramid, are very serious, while at the bottom, are ones which they may choose to ignore.”