The former Manchester United prospect says ChatGPT helped him secure his move to Leyton Orient, raising bigger questions about money, power, and risk in the lower leagues.
Demetri Mitchell did something that sounds like a gimmick and reads like an audit. The 28-year-old used ChatGPT to help negotiate his move to Leyton Orient, structuring counteroffers and arguments before he signed. “They [Leyton Orient] sent me an offer, and I started using ChatGPT, asking it how to negotiate a deal, and what to say in it,” Mitchell said on the From My Left podcast.
Mitchell’s account is granular rather than glossy. He plugged in details about last season’s pay, the cost of moving to London, and family considerations. Then he pushed back. “And then, also because I didn’t use the agent, I get that [agent fee] as a signing-on fee.” The message is simple: keep the portion that would have gone in fees.
This is not a star gaming a superclub. Mitchell is a Manchester United academy graduate who made one league appearance at Old Trafford before spells with Hearts, Blackpool, Hibernian, and Exeter City. He left Exeter on a free this summer and joined Orient without an intermediary, a rare disclosure in a trade that prefers shadows.
He later amplified the story online, framing the arithmetic in plain language and contrasting a typical 5 percent commission with the cost of a subscription that runs about $19 a month. Regardless of where one lands on AI, the post distilled the appeal for players on modest contracts.
What his choice reveals about the lower-league market
Mitchell’s decision is important because it matches how things work in League One. Contracts are usually short, and wages are often lower than fans expect. When improving a deal, the changes are often in the hundreds, not tens of thousands of pounds. If an agent adds a small increase to the salary while still taking a commission, it can make self-representation seem like the best option for the player. His approach also hints at a new layer of professionalism among players, treating negotiation as prep work rather than brinkmanship, with a chatbot acting as a tireless coach for tone, structure, and timing.
There is also a clear regulatory runway. The FA’s Football Agent Regulations state that any clause that limits a player’s ability to negotiate an employment contract on their own, or penalizes them for doing so, “is not permitted and will be null and void.” In other words, self-representation is allowed and protected.
That does not make AI a universal agent. The fine print still matters. Image rights, appearance fees, promotion and relegation wage adjustments, injury protections, and early-termination triggers are exactly where expertise earns its keep. Several sports-law analyses stress how the agent rulebook has grown more complex since FIFA and the FA rewrote it, which is a polite way of saying a lawyer’s read is still prudent even when the player leads talks.
The response highlights a growing tension in the discussion. The AI opportunity, albeit at the risk of misguided “slop,” may allow lower-league players to recover lost earnings. However, there are concerns that prompt culture weakens authenticity because players, like robots, resort to template-based communications. The optimistic view says tools like ChatGPT are becoming negotiation coaches, not replacements, and that confidence rises when players can rehearse their case. The skeptical view thinks you can’t automate the handshake.
In Mitchell’s telling, the club saw the person, not the process. He comes across as prepared, clear about his needs, and realistic about the scale of the deal. It’s more spreadsheet than spectacle. It also feels like a preview. If a League One player can keep the portion that would have gone in fees by steering talks himself, others will try to do the same.
Right now, Mitchell’s experiment is a small case study showing that AI’s true effect will be seen in the details where most football careers actually take place.
