(Dusseldorf) With the world sporting scene at a standstill, the WorldAnti-DopingAgency(WADA) has s’estturnedtoartificial intelligence to detectathleteswho cheat.
WADA is currently funding four research projects in Canada and Germany to see if artificial intelligence (AI) can detect signs of drug use that could escape human investigators. Research is also trying to identify all ethical questions related to this technology.
l’IA Athletes will not be suspended solely on AI testimony. Instead, this technology will be used to ensure that these suspect athletes are tested.
“When you work for an anti-doping organization and you want to target athletes, you look at their competition schedule, their travels and their previous results,” said WADA Senior Executive Director Olivier Rabin. But the human brain has a limit to what it can analyze. »
The pandemic has forced the suspension of anti-doping procedures in several countries, which has put AI’s work at the forefront, as most searches can be conducted remotely.
Analyzing blood or urine samples is more than just finding doping products. Tests can track many biological markers, such as the level of red cells or testosterone.
This type of information is already used by anti-doping agencies through the biological passport, which detects the effects of the use of products such as EPO for example.
WADA hopes that artificial intelligence will help improve this program by following biological markers that can be cross-referenced with other information.. One of WADA’s projects aims to make THE detection of EPO more accurate, while another hopes to do the same with steroids.
Dr.Rabin also spoke of a more “global” project in Montreal, which could predict the risk of doping by evaluating data from a larger number of sources, possibly location data that athletes must provide. The personal data of the athletes and the names of the cities where they reside will be depersonalized to preserve their privacy.
“The discussions around these projects are very complex. […] We need to strike a balance between data protection, protecting individuals while being able to reveal the potential of AI, if there is potential,” Rabin noted.
The results of athletes competing are not yet part of the equation.
“Maybe in the future, but not for now,” Rabin admitted.
Artificial intelligence can be a very expensive science and its specialists are in great demand. The three projects in Canada cost WADA approximately US$425,000 over two years. The Quebec government’s research fund will spend the same amount. The EPO detection project in Germany will cost WADA $60,000.