Charles Woodson spent 18 years in the NFL studying film, preparing for opponents, and making sure he knew exactly what he was talking about before he stepped on the field. Now he’s an analyst on Fox NFL Kickoff, where preparation matters just as much.
So it’s worth noting that, before the start of this season, Woodson asked ChatGPT what the Packers’ division record was last year, got an incorrect answer, and repeated it on national TV anyway.
The AI told him Green Bay went 0-6 in the NFC North. The actual record was 1-5.
“This is the age of AI, and we use it for everything,” Charles Woodson said. “What I learned this year is that the first week of the season, I went on ChatGPT and I just asked a question. I said ‘How many games in the division did the Packers win last year?’ And ChatGPT said they were 0-6 in the division. So, I thought it was an easy enough question. I come up here, I said it with my chest, I’m talking about the Packers. I’m like, ‘Man, they didn’t win a game in the division last year.’ As soon as I got on my [X], somebody quickly corrected me. They said, ‘Hey man, you know the Packers won one division game last year. They beat Chicago.’ I said, ‘Ahhhh, man.’ So, make sure, man, if you’re using AI, that you fact-check.”
😂 @CharlesWoodson learned a valuable lesson to double check ChatGPT pic.twitter.com/DgHT2nEzLl
— FOX Sports: NFL (@NFLonFOX) January 4, 2026
Woodson learned his lesson. Good for him. But the bigger question is why a Hall of Fame defensive back who works for one of the sport’s biggest broadcasters is asking ChatGPT for basic statistics in the first place. Division records aren’t obscure data points. The Packers went 1-5 in the NFC North in 2024, with their only division win coming against the Bears. That’s information you’d find by glancing at the standings or scrolling through last season’s results for five minutes.
ChatGPT was one win off, which isn’t exactly catastrophic but still raises questions about what data it was even pulling from. The technology operates by predicting what words should come next based on patterns in its training data. It doesn’t know anything. It can’t fact-check itself. It can’t distinguish between accurate information and something that sounds plausible. When it’s wrong, it delivers that wrong answer with the same confidence it brings to correct responses.
Sports media has already seen what happens when companies try to cut corners with AI. Sports Illustrated published articles written by AI with fake author bylines and AI-generated headshots. Gannett’s high school sports experiment produced coverage filled with errors and awkward phrasing, including placeholders that made it to publication. These weren’t experiments that failed because the technology wasn’t quite ready. They failed because the companies tried to replace human journalists with AI systems that couldn’t do the job properly.
Charles Woodson’s mistake fits into this pattern, even if it’s smaller in scale. He wasn’t trying to replace journalism with AI. He just wanted a quick answer to what seemed like a straightforward question. But ChatGPT couldn’t even handle that. If AI can’t reliably tell you a team’s division record from the previous season, how can it be trusted with anything more complex?
The appeal of these tools is obvious. They’re fast, they’re convenient, and they give answers with confidence. That confidence is the problem. ChatGPT doesn’t qualify its responses with uncertainty or tell you when it’s guessing. It just gives you an answer that sounds authoritative. For someone trying to quickly prepare for a broadcast, that’s tempting. Why spend time digging through stats when you can just ask a chatbot?
Because these systems get things wrong constantly, and they do it with complete confidence. ChatGPT doesn’t hedge or tell you when it’s guessing. It just gives you an answer. Sometimes that information is accurate. Sometimes it tells you the Packers went 0-6 when they actually went 1-5. You won’t know which you’re getting without checking it yourself, which defeats the entire purpose of using the tool.
The sports media industry should be paying attention. As these tools become more widespread, the temptation to rely on them will only grow. Companies are already experimenting with AI-generated content — including Fox — and plenty of them have gotten burned.
Whether the rest of the industry learns from that remains to be seen.
About Sam Neumann
Since the beginning of 2023, Sam has been a staff writer for Awful Announcing and The Comeback. A 2021 graduate of Temple University, Sam is a Charlotte native, who currently calls Greenville, South Carolina his home. He also has a love/hate relationship with the New York Mets and Jets.