MLSE's AI Experiment at the Deadline Reveals a Fractured Front Office
The Toronto Maple Leafs' trade deadline on March 6 was already a chaotic and largely fruitless affair. But a new report from The Athletic's Chris Johnston has added another layer to the dysfunction; one that raises serious questions about how this organization actually operates at its highest levels.
According to Johnston, MLSE CEO Keith Pelley showed up to the deadline not as a passive observer, but as an active participant; one who came armed with AI-generated trade return scenarios.
"As one team source put it, fans would have been surprised to see just how involved a non-hockey executive was that day," Johnston wrote. "Pelley didn't come empty-handed either. He had notes that included possible trade returns that Leafs staff members believed were generated by Large Language Models and Artificial Intelligence — tools that have become increasingly influential at the top of MLSE but hadn't been part of the usual process with its NHL team."
Johnston also noted that Pelley's musings with AI came from direct orders from Humza Teherany, MLSE's chief strategy and innovation officer. While Teherany was not present that day with the Leafs, his influence loomed large due to him being one of Pelley's most trusted advisors.
This directly contradicts what Pelley said publicly. During his media availability on March 31, he characterized himself as a bystander during the deadline. The reporting tells a very different story; that he was in the thick of it, floating trade ideas and pushing for the Leafs to extract more assets from deals.
For a CEO with no prior hockey operations experience to be that hands-on during one of the most consequential days of the NHL calendar is jarring on its own. What makes it worse is that it unfolded while then-GM Brad Treliving was already fighting an uphill battle trying to move Matthew Knies and squeeze anything of value out of two rental players the Leafs couldn't afford to keep.
"Most vexing, even shocking, for his front office was how little interest rivals seemed to have in Laughton or McMann, two low-cost rentals who'd priced their way out of staying in Toronto. With only 30 minutes to go, Treliving emailed the NHL's other 31 GMs to inform them the Leafs were trading Laughton," Johnston continued. Pelley would later tell MLSE staff in his weekly "Pelley's Page" internal newsletter that the team hadn't drawn any offers on either McMann and Laughton with just 15 minutes to go until the 3 p.m. ET deadline."
Johnston also reported that opposing executives found the Leafs frustrating to negotiate with, describing the calls as more like exploratory conversations than genuine trade discussions. Toronto came close on several occasions; including a reported pursuit of Yegor Chinakhov; but couldn't get anything across the finish line. The only roster move Treliving ultimately made was claiming Troy Stecher off waivers.
The cumulative weight of that dysfunction eventually cost Treliving his job. He was relieved of his duties on March 30, and Craig Berube's future behind the bench remains uncertain depending on who MLSE brings in as the next head of hockey operations. There was no single turning point that sent a team which reached Game 7 of the second round less than a year ago spiraling toward a bottom-five finish; it was a cascade of missteps that compounded month after month.
The revelation that Pelley brought AI-generated analysis into the war room at crunch time offers a window into why he reportedly favors a data-centric candidate for the GM role. It also paints a picture of an organization where the lines between hockey operations and corporate leadership had blurred in ways that created friction rather than clarity. At a moment when the Leafs needed decisive, experienced hockey minds working in concert, they had something considerably messier.
Jeff O'Neill didn't hold back on Overdrive when the story broke, and his reaction cut to the heart of why this matters beyond just one chaotic deadline day.
"I actually think the brand is taking a hit," O'Neill said plainly. He pointed out that the Maple Leafs used to carry a weight that transcended the sport in Canada; that being a Leaf fan was less a choice than a birthright. He argued that era is eroding, and stories like this one accelerate that erosion on multiple levels; from the fan base all the way into the organization itself.
O'Neill went further when addressing how a star player would receive this news.
"If I was a great player and I found out the guy running the hockey operation had some kind of weird scientist that he had no idea was in the war room running the operation; I'd be like, what the hell's going on here?" he said. "I want to buy into this and believe in this. We have R2-D2 in the war room."
That last line got a laugh, but the point underneath it was serious. Players at the top of the game want to feel like hockey people are making hockey decisions. When the process leaks out and it involves AI-generated trade return scenarios being floated by a corporate executive with no hockey background, that is not a culture a franchise player signs up to trust. O'Neill summed it up bluntly; "Hockey people should be doing hockey work, and that's just the way it is."
The GM search itself has done little to quiet concerns about the direction MLSE wants to take this franchise. The name that keeps surfacing is Suhail Deo, but perhaps no candidate has generated more chatter than Sunny Mehta; a man whose background reads less like a traditional hockey executive and more like a case study in unconventional hiring. Mehta won Stanley Cups with the Florida Panthers, which should be the headline. Instead, the conversation around him consistently gravitates toward his scientific background and his former career as a professional poker player.
That tells you something. When the most talked-about quality of a GM candidate is not what he won or how he built a roster but rather what he did before he ever set foot in a hockey front office, it suggests the people driving the search are more excited about the profile than the track record. For an organization that just had a non-hockey executive walk into its war room with AI printouts, the appetite for a data-first, analytically driven candidate makes sense internally. Whether it makes sense for a locker room that needs rebuilding from the culture up is a different question entirely.
Auston Matthews is watching this process unfold in real time. So is every free agent and trade target the Leafs will pursue this summer. The brand, as O'Neill put it, is already taking a hit. Doubling down on the same data-obsessed philosophy that contributed to this mess; just with a shinier hire attached to it; may not be the reset this franchise actually needs.
Post a Comment