Bots that click on advertisements and create fake volumes of data are constantly evolving and while AI technology may help decrease the problem, it won’t disappear entirely, experts said at a panel.

The strategy lead of cybersecurity firm White Ops, Jay Benach, said his company first developed a technology back in 2013 that detected these bots, but advertisers wanted a way that would stop them from making it into the ecosystem in the first place.

The company then developed a machine learning algorithm that did this, and six years later many advertisers think the problem doesn’t exist anymore and fraud is under control. However this is absolutely not the case, proven by the fact that the US government orchestrated a takedown of the “largest ad fraud bot ever”.

“it was operating throughout the entire advertising ecosystem for years.” In Europe and the US and was “baked into advertising models”, said Benach.

The fraud infected 1.7 million machines and mimicked human behaviour, wasting advertising dollars and helping criminals behind it make a lot of money.

Benach warned an over-reliance on data models lead to errors because fraud is not always a clear anomaly, adding criminals need to only trick a small part of the ecosystem to make huge profits

James Hilton, global CEO at marketing agency M&C Saatchi Performance said hackers “use devious ways to keep pushing and maximising revenue” through bots and “by spotting them through AI we can be a step ahead”.

Hilton urged the industry to work together to get the problem under control, while admitting it is naïve to believe it can ever be eliminated completely.