ChatGPT Has Struggled With NFL Betting Picks
The AI model has finished with a 2-8 record through two weeks of play compared to a human rival.
Updated on Sept. 27, 2026 in Football
Live Poll
Do you trust artificial intelligence to make better betting decisions than a human expert?
ChatGPT has underperformed against a human handicapper in an NFL betting challenge, logging a 2-8 record through the first two weeks of the season. The AI model utilized GPT-6 Pro powered by GPT-6 Astra to generate its predictions.
Why it matters
The experiment highlights the limitations of using advanced language models for complex probability-based tasks like sports gambling. Despite consulting ten sources per game, the AI has failed to outperform a human who manages risks more effectively.
ChatGPT holds a 2-8 betting record and a loss of 6.50 units, compared to a 5-5 record and 0.24-unit loss for the human participant. The AI model has a 0-7 record on point spreads and moneylines across the first two weeks of the NFL season.
The players
ChatGPT
This is an artificial intelligence model developed by OpenAI that is currently being tested for its ability to analyze sports betting markets.
NFL
The National Football League is the premier professional football organization in the United States.
The details
Both competitors select five best bets each week using identical odds, with the AI taking 30 minutes to generate its selections after reviewing ten sources per game. While the AI remains winless on spreads and moneylines, the human competitor has successfully backed Cleveland and correctly predicted an under in the Panthers-Falcons matchup.
Timeline
Week 1 saw ChatGPT finish with an 0-5 betting record.
Week 2 resulted in a 2-3 record for the AI model.
September 26, 2026, was when the AI model received the latest odds board at 6 p.m. ET.
Week 3 marks the upcoming attempt for the AI to secure its first winning week.
Season Trajectory
The poor performance of GPT-6 Pro powered by GPT-6 Astra relative to a human handicapper challenges the notion that massive data aggregation provides a superior edge in sports wagering. This outcome may force a reevaluation of how betting markets integrate automated intelligence models.
The AI is currently projecting a 24-17 Bengals win against the Steelers and a final score of Rams 23, Broncos 20. These projections reflect the models attempt to overcome its 0-7 record on point spreads and moneylines.
The takeaway
Reliant on data synthesis, the AI model has struggled to account for the nuance required in professional sports betting. Gamblers should note that algorithmic speed does not currently substitute for the strategic depth demonstrated by human handicappers.
What happens next
ChatGPT is scheduled to attempt to record its first winning week during Week 3 of the NFL season.
Further reading
For more on the current landscape of the professional game, visit the Football section.
Source note: This article includes information reported by The Beltway Report.
Live Poll
Do you trust artificial intelligence to make better betting decisions than a human expert?
