Anti-doping system grows leaps and bounds due to data and AI
The director general of the International Testing Agency (ITA), Benjamin Cohen, says that “without data, they cannot stop anything” in the fight against doping practices in Paris 2024.
In recent years, the fight against doping in sport has been facing a new challenge: the sophistication of doping techniques and the increasing amount of data available. To combat this problem, the International Testing Agency (ITA) is turning to artificial intelligence (AI) as a key tool.
The Swiss jurist, Benjamin Cohen, director of the ITA, the body that will manage the Olympic programme for the third time, after the Tokyo and Beijing Games, says the main task is to fine-tune the “risk analysis”. By this, he means to propose a specific and efficient follow-up of athletes whose previous results have seemed suspicious.
Deep and comprehensive work
“In the last 25 years, the anti-doping system has been greatly strengthened and the possibility of an athlete doping has been drastically reduced,” says Cohen. Aware of modern methods, where genetic and technological doping -which science allows to develop- have come into play, the director is calm, revealing that the conservation of the samples for ten years allows them to go back and analyze each case carefully.
When questioned about the modus operandi of the ITA, he revealed that “the discipline is analyzed” – its physiology, its pharmacology, the most common substances; then, the delegation -the country’s relationship with doping, possible violations of the anti-doping code-; and finally, the athlete is put under the microscope through the evolution of his results, the suspicious profiles of his blood passports… etc.
“There are hundreds of thousands of pieces of data collected to focus on the athletes and the highest risk disciplines,” Cohen concludes on the subject of the investigation.
“Performance passport”, a key ally thanks to AI
Naturally, the task is not entirely simple. On the eve of the ceremony, there are still 30,000 potentially classifiable athletes who will join the more than 11,000 definitive for Paris 2024. In this paradigm, it is vital to optimize the treatment of a large amount of biological data and sports results, as well as the ITA’s anti-doping tests.
“Without data, we cannot stop anything,” he admitted, and for this reason, “we work on them constantly”, Cohen concluded bluntly. The improvements do not stop, and since the Tokyo Games held in 2021, the ITA has developed “a performance passport” to measure and detect extraordinary improvements in performance.
Cohen also warned that the ITA will “predict some results based on the performance of athletes four years in advance.” This is the work of the phenomenon of artificial intelligence, whose rise has been meteoric, and on this occasion, it will allow for the qualification of an unusual result that could perhaps be due to a case of doping.
The Swiss lawyer thus reports that the project has been tested in swimming and weightlifting, and recently presented to the cycling family: “It is a tool that we would like to develop in this discipline, whether on the track or against the clock, thus being able to measure individual power and performance,” he concludes.
“More progress” with a view to Los Angeles 2028
“Not long ago, everyone worked with an Excel file,” Cohen recalls about the routine tasks that were carried out some Games ago. However, they now have their “own program” which still has room for improvement, since “there is still a lot of manual work,” he adds.
According to Cohen’s speech, the ITA’s challenges include creating an interactive database that can be used to compare data with other organisations and to jointly promote the fight against doping in future events.
When questioned about the upcoming games in California, the Swiss director said that they will have made much more progress in terms of technology, in terms of research.
However, he focuses on logistics, “the most complicated point” in his opinion, predicting that in four years’ time, “more innovative collection methods, including drones to go faster.”