ESPN's AI Fumble: How Robot Recaps Missed Alex Morgan's Final Match
ESPN's AI Fumble: How Robot Recaps Missed Alex Morgan's Final Match
As ESPN turns to artificial intelligence for game coverage, a soccer legend's farewell exposes the technology's limitations.
Sports media giant ESPN recently announced a major shift in its coverage strategy. Starting this month, the network began using AI to generate game recaps for two leagues: the National Women's Soccer League (NWSL) and the Premier Lacrosse League (PLL). ESPN frames this move as an effort to expand coverage of "under-served sports," arguing that AI-generated content is better than no coverage at all.
The limitations of this approach became glaringly apparent this weekend. On Sunday, U.S. soccer legend Alex Morgan played her final professional match, a 4-1 loss by her San Diego Wave to the North Carolina Courage. The game featured an emotional farewell, with Morgan removing her cleats at midfield and addressing the crowd. Yet, as the X (formerly Twitter) account Sports TV News & Updates noted, the AI-generated recap, supposedly reviewed by human editors, made no mention of Morgan or the significance of the match entirely. This glaring omission raises serious questions about the ability of AI to capture the context and emotional weight of newsworthy events.
The next morning, the article was updated to include a short paragraph about Morgan’s farewell, referring to a failed penalty kick attempt as a shot on goal:
It was the final game in the nearly 14-year career of US star Alex Morgan. The two-time World Cup winner and Olympic gold medalist played 15 minutes, exiting in the first half. Her shot on goal in the 10th minute was saved by Courage goalie Casey Murphy.
Writing about the flub at Awful Announcing, Sean Keeley asked an important question: “If you’re going to provide fans of underserved sports with poorly written recaps that don’t tell the full story, what value are you really providing them?”
ESPN's decision to use AI for game recaps has sparked debate and garnered criticism from within the journalism community. As Tom Jones notes in a piece for Poynter, the move raises questions about the future of sports journalism and the role of AI in media:
"As with anything involving the rapidly changing alliance of AI and journalism, what followed were lots of questions and a little uneasiness."
ESPN argues that this approach allows them to cover leagues that currently lack full coverage by human journalists. In their announcement, the network stated:
"The AI-generated recaps aim to enhance coverage of under-served sports, providing fans with content that was previously unavailable. These sports do not currently have game recaps on ESPN digital platforms, and these AI-generated recaps will be a tool to augment existing coverage — not replace it."
However, critics argue that this approach may be a slippery slope. Jones relays this concern:
"The fear among living and breathing journalists [is] that this is a slippery slope, and that AI is taking their jobs. They say: Instead of using AI to write lacrosse and soccer stories, why doesn't ESPN just hire more journalists? And if ESPN can use AI to cover those sports, what's to stop them from using AI to cover more and more other sports?"
ESPN Chairman Jimmy Pitaro has defended the move, stating at a recent media day:
"AI is not the awful, terrible disruptor that many people think it will be. I personally believe that AI will be very helpful and beneficial to the sports industry and the sports fan, and we're starting to see that."
This optimistic view, however, is challenged by incidents like the Morgan oversight, which illustrate the limitations of AI in capturing the nuances and significance of sporting events.
As AI continues to make inroads into journalism, the Morgan incident serves as a cautionary tale. While AI may offer a way to expand coverage, it's clear that the technology still struggles with nuance, context, and the human elements that make sports compelling. As media organizations navigate this new landscape, they’ll surely be looking to balance innovation with the core tenets of quality journalism or risk losing readers. This is an ethical minefield.
What I’m reading
“Second Circuit Says Libraries Disincentivize Authors To Write Books By Lending Them For Free” (TechDirt, Mike Masnick, 9/5/24)
“OffBall launches as new sports culture brand” (Semafor, Ben Smith, 9/8/24)
“Musk sticks up for his rightwing buddies, not free speech” (
, , 9/9/24)“The Trump-Musk alliance is about their unquenchable thirst for power” (
, , 9/8/24)“Did Frank Sinatra Really Perform at My Grandma's High School?” (
, , 9/5/24)“The power of a single word about media malfeasance” (
, , 9/7/24)
This approach is not better than nothing, it’s worse than nothing. Because now this bad article becomes part of the Google/Bing/etc. search ecosystem. ESPN is a trusted source, so it could even be cited on Wikipedia. Until it can be trusted to be as good as humans (if that ever happens), it shouldn’t be deployed.
How does AI work in this context? Is an artificial intelligence "watching" the match live and reporting on it, or is AI searching and summarizing (i.e., stealing) content created by humans who watch the match and post content about it on the internet? If AI is stealing, it's both an ethical problem and a sign that AI isn't very I yet. For underserved sports like women's soccer, there are many excellent commenters, podcasters, and writers out there providing excellent content--Sam Mewis, Tobin Heath, and Christen Press to name three whom I follow. They are not formally considered journalists (probably?), but couldn't ESPN pay one or more of them or others in the space as contractors to provide content that would actually be good and valuable to ESPN's subscribers and followers? Even at $1,000/hour, ESPN could get great match summaries that might drive traffic to their site instead of alienating fans, and for a pittance compared to their overall budget.