We asked five AI chatbots to predict the 2026 NFL Draft. Here's how it went

AI may have many uses. Predicting the NFL draft may not be one of them.

As the 2026 NFL Draft approaches, draft experts around the internet have tried to predict the outcome of the first round in their own unique mock drafts. Each expert has provided their own spin on the first 32 picks in this year's draft, and most of them have provided analysis while explaining their thought process for those picks.

But what if we pawned off all of that lousy "thinking" and "analysis" to a large language model (LLM) powered by AI? Can the chatbots come up with their own versions of a mock draft that compete with the accuracy and predictive analysis provided by real draft experts?

AI VS. EXPERT: How chatbot picks compare to expert predictions in 2026 NFL mock draft

Over the last month, USA TODAY Sports has consulted with five different AI chatbots, asking each one to come up with its own predictions for the first round of the 2026 NFL Draft. OpenAI's ChatGPT, xAI's Grok, Microsoft Copilot, Meta AI and Google Gemini each put together its own mock draft.

Here is a full roundup of the results of AI mock drafts ahead of the 2026 NFL Draft:

AI mock draft predictions before 2026 NFL Draft

Note: All five mock draft predictions were compiled before the Cincinnati Bengals traded the No. 10 overall pick to the New York Giants.

1. Las Vegas RaidersMicrosoft Copilot: QB Fernando Mendoza, IndianaGoogle Gemini: QB Fernando Mendoza, IndianaGrok: QB Fernando Mendoza, IndianaMeta AI: QB Fernando Mendoza, IndianaChatGPT: QB Fernando Mendoza, Indiana

The only consensus pick of the entire mock draft roundup. Mendoza to the Raiders is the closest thing to a foregone conclusion in this draft, so even the artificial intelligence couldn't screw this pick up.

2. New York JetsMicrosoft Copilot: RB Jeremiyah Love, Notre DameGoogle Gemini: LB/Edge Arvell Reese, Ohio StateGrok: LB/Edge Arvell Reese, Ohio StateMeta AI: LB/Edge Arvell Reese, Ohio StateChatGPT: QB Dante Moore, Oregon

The Jets' selection at No. 2 was one of the picks that came the closest to a consensus in this roundup of AI mock drafts. One has to wonder how much of that came down to three out of the five chatbots seemingly forgetting about Texas Tech edge rusher David Bailey's existence. Copilot's decision to have Love match Eagles running back Saquon Barkley's draft position at No. 2 overall was a real head-scratcher, but it's not even the worst pick of the group. ChatGPT's selection of Moore, who has already announced he would return to Oregon instead of declaring for the draft, was just a bit of foreshadowing for the disastrous mock draft OpenAI's LLM put together.

3. Arizona CardinalsMicrosoft Copilot: Edge Rueben Bain Jr., Miami (Fla.)Google Gemini: Edge David Bailey, Texas TechGrok: OT Francis Mauigoa, Miami (Fla.)Meta AI: OT Francis Mauigoa, Miami (Fla.)ChatGPT: S Caleb Downs, Ohio State

Spoiler alert: Gemini's pick is the highest that Bailey goes in any AI mock draft in this roundup. Meanwhile, Copilot had Arizona opting for a different edge rusher - Microsoft's LLM had Bailey go later - and the two social media-owned AIs were in lockstep on an offensive tackle selection instead. Bailey and Mauigoa are among consensus selections for the Cardinals in the draft community, so those picks don't stand out. Bain at No. 3 is a bit of a reach in the post-NFL combine world where his arm length raised concerns. Elsewhere, ChatGPT continued to march to the beat of its own drum.

4. Tennessee TitansMicrosoft Copilot: LB/Edge Arvell Reese, Ohio StateGoogle Gemini: RB Jeremiyah Love, Notre DameGrok: RB Jeremiyah Love, Notre DameMeta AI: RB Jeremiyah Love, Notre DameChatGPT: OT Kadyn Proctor, Alabama

Another near-consensus pick. Tennessee would love to have Reese fall to them as one of the top edge rushers in the class, and Love is otherwise a popular consensus pick for the Titans among experts' mock drafts. For ChatGPT, none of that meant anything. While building up the offensive line in front of Cam Ward makes sense, no draft expert has Proctor as the top offensive tackle in this class, which is where he lands in OpenAI's mock.

5. New York GiantsMicrosoft Copilot: S Caleb Downs, Ohio StateGoogle Gemini: LB Sonny Styles, Ohio StateGrok: LB Sonny Styles, Ohio StateMeta AI: LB Sonny Styles, Ohio StateChatGPT: WR Carnell Tate, Ohio State

According to five different AI large language models, one thing was for sure when it came to the No. 5 pick: an Ohio State Buckeye was coming off of the board. Though the LLMs did not have a consensus pick, all of the selections made sense for New York. Downs and Styles provide a necessary upgrade to Big Blue's defense, while Tate is an excellent wide receiver to pair with lead wideout Malik Nabers to support quarterback Jaxson Dart.

6. Cleveland BrownsMicrosoft Copilot: OT Francis Mauigoa, Miami (Fla.)Google Gemini: OT Francis Mauigoa, Miami (Fla.)Grok: OT Monroe Freeling, GeorgiaMeta AI: OT Spencer Fano, UtahChatGPT: DT Peter Woods, Clemson

Build up the trenches! Getting an offensive tackle in the building is important for protecting whichever quarterback wins the starting battle in Cleveland for 2026, whether that be Mauigoa, Fano or Freeling (though the latter is a bit of a reach at 6). ChatGPT continues to do its own thing here, and while Woods is also a reach this early, he is also one of the top defensive tackle prospects in the class.

7. Washington CommandersMicrosoft Copilot: LB Sonny Styles, Ohio StateGoogle Gemini: WR Carnell Tate, Ohio StateGrok: S Caleb Downs, Ohio StateMeta AI: Edge Rueben Bain Jr., Miami (Fla.)ChatGPT: Edge Rueben Bain Jr., Miami (Fla.)

The Commanders' pick was another rare instance in which there were several different options from the different LLMs and all of them would make sense. Styles is a pick that would make a ton of sense for Washington if he fell this far - helping to replace Bobby Wagner in the linebackers group and provide a capable pass defender at the position. Downs would help give the Commanders an extra boost in the secondary, which they desperately need with a weak group of cornerbacks. Bain could slot in across from free agent signing Odafe Oweh to upgrade Washington's pass-rush attack in a big way, and Tate could be the No. 2 wide receiver the Commanders are in need of behind Terry McLaurin.

8. New Orleans SaintsMicrosoft Copilot: WR Carnell Tate, Ohio StateGoogle Gemini: CB Mansoor Delane, LSUGrok: WR Carnell Tate, Ohio StateMeta AI: CB Mansoor Delane, LSUChatGPT: WR Evan Stewart, Oregon

Addressing the worst pick first: ChatGPT's selection of Stewart was the second time in its mock draft it tried to pick a player that did not declare for the draft. Outside of that, it was either Tate - to give New Orleans quarterback Tyler Shough a strong second option to throw to outside of Chris Olave - or Delane, a cornerback that fills a need across from Kool-Aid McKinstry after Alontae Taylor's departure in free agency. Not bad.

9. Kansas City ChiefsMicrosoft Copilot: WR Jordyn Tyson, Arizona StateGoogle Gemini: S Caleb Downs, Ohio StateGrok: Edge Rueben Bain Jr., Miami (Fla.)Meta AI: WR Jordyn Tyson, Arizona StateChatGPT: OT Will Campbell, LSU

Giving Mahomes more wide receiver help with Tyson would make sense at No. 9, as would improving the secondary or edge rusher situation in Kansas City. The Chiefs have not done well getting after opposing quarterbacks on the edge - though Chris Jones has held things down in the interior - and they've lost some key defenders in the secondary this offseason.

What does not make sense in the slightest is wasting a top-10 pick on a player that already plays for another professional team. This is only the beginning of many (many) such moves by ChatGPT.

10. Cincinnati BengalsMicrosoft Copilot: DL Peter Woods, ClemsonGoogle Gemini: Edge Rueben Bain Jr., Miami (Fla.)Grok: LB Sonny Styles, Ohio StateMeta AI: S Caleb Downs, Ohio StateChatGPT: OT Kelvin Banks Jr., Texas

Improving the defense had to be a clear point of focus for the Bengals this offseason - which they made clear in their trade of this pick for Dexter Lawrence. These AI mock drafts were all put together before that trade, and each LLM understood that assignment except for ChatGPT. Offensive tackle is a position of need, but Cincinnati will have a hard time if it tries to draft one of the Saints' players.

2026 NFL DRAFT: Small-school standouts led by San Diego State CB

11. Miami DolphinsMicrosoft Copilot: Edge David Bailey, Texas TechGoogle Gemini: WR Makai Lemon, Southern CaliforniaGrok: WR Makai Lemon, Southern CaliforniaMeta AI: WR Makai Lemon, Southern CaliforniaChatGPT: CB Will Johnson, Michigan

Lemon's a good fit for a team that released wide receiver Tyreek Hill and traded away Jaylen Waddle this offseason. Bailey is a steal at the No. 11 pick in the draft and gets the Dolphins a high-ceiling replacement for Bradley Chubb, whom they also released in the offseason. But once again, ChatGPT is trying to use the draft to steal players from other teams. Johnson was the Cardinals' second-round pick last year.

12. Dallas CowboysMicrosoft Copilot: CB Mansoor Delane, LSUGoogle Gemini: CB Jermod McCoy, TennesseeGrok: LB Sonny Styles, Ohio StateMeta AI: CB Jermod McCoy, TennesseeChatGPT: Edge James Pearce Jr., Tennessee

Defense, defense, defense for the Cowboys, who finished last year with one of the worst units in the sport. Dallas allowed the third-highest completion rate, second-highest passer rating and most passing yards to opponents in 2025, so adding a cornerback makes sense. Styles, a super-versatile freak athlete, would be a steal for Dallas here as well. But not as much of a steal as Pearce, who is currently rostered by the Atlanta Falcons.

13. Los Angeles Rams (from Atlanta Falcons)Microsoft Copilot: OT Spencer Fano, UtahGoogle Gemini: WR Jordyn Tyson, Arizona StateGrok: WR Makai Lemon, Southern CaliforniaMeta AI: TE Kenyon Sadiq, OregonChatGPT: Edge Mykel Williams, Georgia

Fano would give the Rams talented extra depth at offensive tackle, while Tyson and Lemon would do the same at wide receiver to eventually replace Davante Adams. Notably, Grok just re-drafted Lemon two picks after sending him to the Dolphins. Sadiq would be a strong addition to an offense that loved to run three-tight-end sets in 2025, though there isn't an immediate need there. Meanwhile, ChatGPT tries to steal a player who is already in the NFL for a fifth straight pick.

14. Baltimore RavensMicrosoft Copilot: Edge T.J. Parker, ClemsonGoogle Gemini: OG Olaivavega Ioane, Penn StateGrok: OG Olaivavega Ioane, Penn StateMeta AI: OG Olaivavega Ioane, Penn StateChatGPT: CB/WR Travis Hunter, Colorado

Make it six straight stolen players for ChatGPT as Hunter, currently on the Jaguars, apparently manages to both play offense and defense and play for two different teams. Otherwise, Ioane is the most popular pick for a Ravens offensive line that just parted ways with center Tyler Linderbaum. Edge rusher is not as pressing a need for Baltimore after it signed Trey Hendrickson, and the Parker pick is also a bit of a reach by Copilot. Still not as bad as ChatGPT.

15. Tampa Bay BuccaneersMicrosoft Copilot: WR Makai Lemon, Southern CaliforniaGoogle Gemini: Edge Keldric Faulk, AuburnGrok: Edge Akheem Mesidor, Miami (Fla.)Meta AI: Edge Akheem Mesidor, Miami (Fla.)ChatGPT: WR Tetairoa McMillan, Arizona

Replacing wide receiver Mike Evans is on the docket for the Buccaneers in the logic of Copilot and ChatGPT, but only one of them picked a draftable player. McMillan, the reigning Offensive Rookie of the Year for the Carolina Panthers, isn't going anywhere. The three other LLMs opted to address a more pressing draft need for Tampa Bay: edge rusher.

16. New York Jets (from Indianapolis Colts)Microsoft Copilot: QB Ty Simpson, AlabamaGoogle Gemini: QB Ty Simpson, AlabamaGrok: WR Jordyn Tyson, Arizona StateMeta AI: WR Denzel Boston, WashingtonChatGPT: Edge LT Overton, Alabama

A new quarterback for Gang Green? That makes sense, Copilot and Gemini. More wide receiver help across from Garrett Wilson? Solid decisions, Grok and Meta AI. An edge rusher that ranks outside the top 75 of USA TODAY Sports' draft expert Ayrton Ostly's big board? Well, ChatGPT, at least it's a 2026 draft prospect for the first time since No. 7 overall.

17. Detroit LionsMicrosoft Copilot: OT Monroe Freeling, GeorgiaGoogle Gemini: OT Kadyn Proctor, AlabamaGrok: OT Monroe Freeling, GeorgiaMeta AI: OT Kadyn Proctor, AlabamaChatGPT: DT Mason Graham, Michigan

Offensive tackle is a clear need for the Lions with Taylor Decker's surprise release request, and four out of five AI LLMs got that memo. If you had one guess as to which missed it... That's right, ChatGPT picked another player that already plays in the NFL.

18. Minnesota VikingsMicrosoft Copilot: CB Jermod McCoy, TennesseeGoogle Gemini: S Dillon Thieneman, PurdueGrok: S Dillon Thieneman, PurdueMeta AI: S Dillon Thieneman, PurdueChatGPT: OT Aireontae Ersery, Minnesota

Thieneman is by far the most popular pick for the Vikings at No. 18 in draft experts' mocks around the web, so it makes sense to see him well-represented in the LLMs' picks. McCoy also makes some sense for Minnesota as help in the secondary at a different position. What doesn't make sense? Drafting an offensive tackle that already plays for the Houston Texans.

19. Carolina PanthersMicrosoft Copilot: WR Denzel Boston, WashingtonGoogle Gemini: TE Kenyon Sadiq, OregonGrok: WR Makai Lemon, Southern CaliforniaMeta AI: OT Monroe Freeling, GeorgiaChatGPT: TE Colston Loveland, Michigan

More pass-catching help for Bryce Young? Sure, even if it means a third straight year of taking a wide receiver in the first round. For Grok, this is the third time Lemon has been drafted, so that's not ideal. Gemini's Sadiq selection is the best of the bunch here by getting tight end help instead. While ChatGPT's "heart" was in the right place, its logic once again misled it by trying to draft yet another player that has already been employed by another pro team.

20. Dallas Cowboys (from Green Bay Packers)Microsoft Copilot: OT Caleb Lomu, UtahGoogle Gemini: Edge Akheem Mesidor, Miami (Fla.)Grok: Edge Akheem Mesidor, Miami (Fla.)Meta AI: LB CJ Allen, GeorgiaChatGPT: DT Deone Walker, Kentucky

More defensive help for the Cowboys in most of the selections here at No. 20 after most LLMs had them taking a cornerback at 12. Meanwhile, ChatGPT is back to its old ways and trying to give Dallas another already-drafted player: first it was James Pearce Jr., now Walker, who was the Bills' fourth-round pick last year.

NFL TRADE PREDICTIONS: A.J. Brown tops NFL draft trade candidates

21. Pittsburgh SteelersMicrosoft Copilot: DL Caleb Banks, FloridaGoogle Gemini: WR Denzel Boston, WashingtonGrok: OG Olaivavega Ioane, Penn StateMeta AI: Edge T.J. Parker, ClemsonChatGPT: CB Benjamin Morrison, Notre Dame

No real theme emerges from the collective AI picks at No. 21. In addition to ChatGPT continuing its tendency to select 2025 draft picks (Morrison was a second-rounder for the Buccaneers last year), Grok got in on the weirdness. The xAI chatbot tried to select Ioane a second time after sending him to the Ravens seven picks prior.

22. Los Angeles ChargersMicrosoft Copilot: CB Chris Johnson, San Diego StateGoogle Gemini: OT Spencer Fano, UtahGrok: OG Olaivavega Ioane, Penn StateMeta AI: OT Blake Miller, ClemsonChatGPT: OT Josh Conerly Jr., Oregon

So Grok apparently loves re-drafting players in the same draft. First Lemon, now Ioane gets drafted a third time. ChatGPT still loves stealing players from other teams, with the Commanders losing their offensive tackle this time. Tackle depth makes some sense here from Meta AI and Gemini's picks after last year's disaster. Johnson, one of the best small school prospects in this class, gets into the first round in Copilot's mock draft.

23. Philadelphia EaglesMicrosoft Copilot: TE Kenyon Sadiq, OregonGoogle Gemini: S Emmanuel McNeil-Warren, ToledoGrok: TE Kenyon Sadiq, OregonMeta AI: Edge Keldric Faulk, AuburnChatGPT: WR Luther Burden III, Missouri

If the Eagles are able to get Sadiq, their eventual replacement for Dallas Goedert, at No. 23, their draft would be off to an incredible start. One thing Howie Roseman definitely won't do? Try to poach Luther Burden III from the Bears with his first-round pick, at least in any way that isn't via trade. McNeil-Warren and Faulk would address Philly's needs on defense, with edge rusher among the biggest roster holes.

24. Cleveland Browns (from Jacksonville Jaguars)Microsoft Copilot: OT Kadyn Proctor, AlabamaGoogle Gemini: WR Chris Brazzell II, TennesseeGrok: WR Denzel Boston, WashingtonMeta AI: WR Omar Cooper Jr., IndianaChatGPT: Edge Laiatu Latu, UCLA

Disregarding ChatGPT again here, the rest of the picks mostly make sense. Copilot's move to have Cleveland double down on offensive tackles in the first round is questionable, but the wide receiver picks from other LLMs make a ton of sense for a team needing more support in the receiving corps.

25. Chicago BearsMicrosoft Copilot: WR Omar Cooper Jr., IndianaGoogle Gemini: DT Kayden McDonald, Ohio StateGrok: S Emmanuel McNeil-Warren, ToledoMeta AI: S Emmanuel McNeil-Warren, ToledoChatGPT: OT Olu Fashanu, Penn State

Once again, we're going to disregard ChatGPT's shenanigans at this point. The Bears could really use safety help after letting so many of their safeties walk in free agency this offseason. Grok and Meta AI's McNeil-Warren picks address that. Copilot's Omar Cooper Jr. selection serves as Chicago's replacement for DJ Moore, whom the Bears traded to the Bills, but is also a more questionable pick with more pressing needs elsewhere.

26. Buffalo BillsMicrosoft Copilot: S Dillon Thieneman, OregonGoogle Gemini: Edge R Mason Thomas, OklahomaGrok: Edge Cashius Howell, Texas A&MMeta AI: CB Avieon Terrell, ClemsonChatGPT: S Malaki Starks, Georgia

Again, ChatGPT makes a horrible pick of a 2025 draft prospect, and again, we disregard. Gemini and Grok's picks for an edge rusher make the most sense here to address one of Buffalo's biggest needs.

27. San Francisco 49ersMicrosoft Copilot: Edge Keldric Faulk, AuburnGoogle Gemini: OT Caleb Lomu, UtahGrok: Edge T.J. Parker, ClemsonMeta AI: OT Max Iheanachor, Arizona StateChatGPT: DT Brandon Dorlus, Oregon

ChatGPT disregarded. Gemini and Meta AI have the 49ers drafting their possible Trent Williams replacement, while Copilot and Grok have them addressing one of their biggest needs at edge rusher. Outside of ChatGPT, solid choices all around.

28. Houston TexansMicrosoft Copilot: LB Anthony Hill Jr., TexasGoogle Gemini: CB Avieon Terrell, ClemsonGrok: DT Kayden McDonald, Ohio StateMeta AI: Edge Cashius Howell, Texas A&MChatGPT: TE Ja'Tavion Sanders, Texas

All AI LLMs (except one, because of course) went with an interesting strategy of going defense with the Texans' first-round pick. Houston already has one of the strongest defensive units in the sport, and offensive line felt like a bigger priority heading into this draft. The chatbots seemed to disagree here at pick 28.

29. Kansas City Chiefs (from Rams)Microsoft Copilot: WR KC Concepcion, Texas A&MGoogle Gemini: Edge Nic Scourton, Texas A&MGrok: CB Jermod McCoy, TennesseeMeta AI: CB Colton Hood, TennesseeChatGPT: CB Kalen King, Penn State

Similar to the Chiefs' first pick in this mock draft roundup at No. 9, pretty much every pick (except the one bad one, you know which) makes sense. Cornerback needs to be a priority with Trent McDuffie and Jaylen Watson going to the Rams earlier in the offseason, and that's what Grok and Meta AI address. Wide receiver and edge rusher help is necessary as well, which jibe with Copilot and Gemini's selections.

30. Miami Dolphins (from Denver Broncos)Microsoft Copilot: OT Blake Miller, ClemsonGoogle Gemini: CB Colton Hood, TennesseeGrok: CB Avieon Terrell, ClemsonMeta AI: DT Kayden McDonald, Ohio StateChatGPT: OT Amarius Mims, Georgia

Another pick, another nonsensical ChatGPT selection. Outside of that, more draft capital spent on building up the Dolphins' foundation makes sense after each AI took strides to do so with the No. 11 pick. Meta AI's selection stands out as the most unique of this bunch as a player at a non-premium position.

31. New England PatriotsMicrosoft Copilot: QB Garrett Nussmeier, LSUGoogle Gemini: DT Peter Woods, ClemsonGrok: OT Blake Miller, ClemsonMeta AI: LB Jake Golday, CincinnatiChatGPT: CB Kool-Aid McKinstry, Alabama

Copilot got in on the jaw-droppingly bad logic train with this pick. The Patriots are not drafting a quarterback in the first round after Drake Maye was the MVP runner up and led New England back to the Super Bowl. ChatGPT also picked another 2025 prospect, so nothing new there, and Meta AI reached for a linebacker. That leaves Grok, shockingly, with one of the better picks to give Maye some blocking help up front. Gemini's defensive tackle pick - getting Woods, a top-ranked player at his position as well - is the best pick here.

32. Seattle SeahawksMicrosoft Copilot: CB Avieon Terrell, ClemsonGoogle Gemini: WR KC Concepcion, Texas A&MGrok: CB Brandon Cisse, South CarolinaMeta AI: CB Brandon Cisse, South CarolinaChatGPT: Edge Chop Robinson, Penn State

ChatGPT closes out its mock with a whimper. Copilot, Grok and Meta AI all make a logical pick for a cornerback with Riq Woolen departing earlier in the offseason. Gemini's selection of a wide receiver is a bit surprising given Seattle's retention of Rashid Shaheed in free agency, but the extra depth wouldn't hurt for the Seahawks' offense.

This article originally appeared on USA TODAY: Five AI chatbots predict the 2026 NFL Draft: Mock draft roundup