What if AI began to replace agents in professional sports to reach contract agreements? Demetri Mitchell, 28, formerly of Manchester United’s youth academy and European Under-17 champion with England, explained on the podcast ‘From My Left’ how he used ChatGPT to guide his negotiations with London-based League One club Leyton Orient. “An agent charges me 5% commission, ChatGPT? Only £15.” Football agents in England pocketed £409 million (€477 million) in commissions between February 2024 and 2025, a significant amount.

Mitchell, who also played for Hibernian, Blackpool, and Exeter City, was a free agent this summer when he received the offer from Leyton Orient. Instead of hiring an intermediary, he asked the AI for advice, asking it relevant questions to try to find the best possible answer, with the best arguments. “I started using ChatGPT and asked it how to negotiate the contract and what to say. I said things like, ‘This is what I earned last season, I would have to move to London, so what is the cost of living there?’… I knew I could negotiate above what they had offered me, but I didn’t want to say directly, ‘I deserve this much.’”

The midfielder not only managed to improve his conditions, but also received a bonus for signing directly, without intermediaries, something unusual in lower leagues. “ChatGPT has been my best agent to date.”

The reference of Kevin de Bruyne

The case is reminiscent of Kevin De Bruyne, who in 2021 negotiated his renewal with Manchester City without an agent. The Belgian used data on his sporting performance, processed by the company Analytics FC, as a tool to convince the club’s directors. Mitchell took a different path, but with the same underlying idea: to rely on tools other than the traditional agent. The difference is that, in his case, he did so with an AI application accessible to any user.

For players like Mitchell, with shorter contracts and less stable salaries, the possibility of avoiding that 5% commission is particularly attractive. “There is a big difference between someone who earns £2,000 or £3,000 a week in the Championship and a doctor who earns the same, because the doctor has it guaranteed until the age of 65. We only have a one or two-year contract.”

Beyond contracts: other applications of AI

Mitchell’s case opens the door to a larger debate: how far can Artificial Intelligence go in professional sports? In soccer, it is already used in data analysis, physical performance, injury prevention, and talent detection. Clubs use algorithms that identify patterns in players’ workloads, recommend personalized training plans, and even predict the risk of injury before it occurs.

In sports such as basketball and tennis, AI is used for advanced scouting and match strategy, with tools capable of processing thousands of actions in seconds and proposing tactical adjustments. Sports marketing has also benefited, from the creation of automated content to the personalization of the fan experience in stadiums and on digital platforms.

Mitchell’s example not only calls into question the role of agents, but also reflects how AI is beginning to penetrate more sensitive areas of sport, such as the management of athletes’ professional careers based on accurate and impartial data.