The Association of National Advertisers (ANA) recently conducted a study that found that marketers are continuing to lose money to bots, the computer programs that create fraudulent Web traffic by mimicking human Web browsing habits. Despite efforts to inform marketers about the dangers of bots, the level of fraud remains relatively unchanged from the previous year. The authors of the study estimate that companies could lose more than $7 billion globally this year because of it.

ANA, in cooperation with the ad fraud detection firm White Ops, tracked the online ad buys of 49 brands over a two-month period. The brands included names like McDonald’s, Anheuser-Busch and Unilever. The study showed that bot traffic ranged between 3 percent to 37 percent. The previous study from 2014 found a similar range with 2 percent to 22 percent of ad impressions coming from bot traffic.

Some ads are much more likely to attract bot traffic than others. The study found that ads, both video and display ads, that are bought using automated systems are more likely to have fraudulent Web traffic compared to ads that were purchased through human sales. Also, the publishers that use “sourced traffic” from third-party companies were three times as likely to have fraudulent Web traffic.

Bob Liodice, chief executive of ANA, says that the advertising industry has been slow to address the problem. “As an industry, when faced with huge issues, we become paralyzed,” he said in The Wall Street Journal.

The issue continues to grow as digital ad spending increases. Display ad spending and online video ad spending increased 24 and 42 percent respectively last year.