Thursday Jul 30
SPORTOther

AI Scores Shaq's New Dunk League

30JUL
SCORED BY AIOWL AIDUNKMAN

A basketball contest just let a computer decide who wins. DUNKMAN, the new pro dunk league, tracks leap, distance, and power with Owl AI cameras. That data feeds the score directly, no judges needed.

Owl AI's cameras clock jump height, hang time, and force on every attempt. Those readings flow straight into the results during Week 2 of competition.

Twenty-four athletes chase a $500,000 prize across four events in Atlanta. TNT Sports built the broadcast around real-time numbers instead of gut calls. Senior director Daniel Nabors says the readouts make results feel earned.

Officiating keeps quietly shifting from human eyes to sensors across sports. This event is an unlikely place to see it land first.

full brief & sources

⚡ Why this matters

  • AI-driven scoring makes judging feel data-backed instead of subjective.
  • A new sports property builds its entire broadcast around computer vision.
  • Leagues keep testing how much judgment they can hand to sensors.

🔍 What happened

  • DUNKMAN launched July 21, 2026, on TNT Sports and HBO Max.
  • Shaquille O'Neal serves as commissioner; 24 dunkers compete for $500,000.
  • Owl AI uses computer vision to capture vertical leap, distance, and dunk power.
  • Metrics feed live into the Ross XPression graphics system during broadcast.
  • TNT Sports senior director Daniel Nabors says the data legitimizes the judging.
  • The season runs four weekly events, ending with an August 25 championship.

💬 Smart takes

  • Daniel Nabors, TNT Sports senior director: the AI graphics package "helped legitimize the judging process by making the scoring more transparent and data-driven."
  • Skeptic: a computer can measure a dunk's height, but crowd reaction and showmanship still get scored by feel, not sensors.

🧭 Where this goes

  1. Likelyother slam-dunk and trick-sport contests add computer vision scoring within a year.
  2. LikelyDUNKMAN leans harder into stat overlays for its championship broadcast on August 25.
  3. Possibleleagues start publishing raw sensor data so fans can second-guess judges themselves.
  4. Wild Carda major sport tries fully AI-scored judging for a marquee event and fans revolt.

🥄 The Spoon Take

Dunking just got a scoreboard that doesn't argue back. Owl AI turns leap height and power into numbers fans can see live, not a judge's gut feeling. Expect every new sports property to copy this playbook before copying the sport itself.

🤔 Pushback

Vertical leap is easy to sensor. Judging which dunk had more flair is still a human call, AI or not.

·SVG - Inside DUNKMAN's AI booth·Variety - Dunkman premiere details
Thursday Jul 23
SPORTOther

MLB Bans AI From The Dugout

23JUL
+ FIFA TOOBENCHEDCOACH CALLS

Baseball just took the robot out of the dugout. MLB banned iPads feeding coaches live AI calls this week. FIFA barred in-game AI days later, drawing sports' first real AI line.

The trigger: MLB found teams using dugout iPads to auto-suggest pitching changes and lineup moves, not just video review.

One executive said a third of the league had built custom AI logic into the tablets. The Mets reportedly led the push.

FIFA's parallel ban means two of the world's biggest leagues now treat live-game AI as a threat to human judgment, not just an edge.

full brief & sources

⚡ Why this matters

  • Live AI in the dugout blurs who's actually managing the game, the coach or the model.
  • Two major leagues drawing the same line in one week signals real consensus, not one outlier.
  • It's the first concrete AI-in-the-loop rule for pro sports with enforcement, not just a policy statement.

🔍 What happened

  • MLB EVP Morgan Sword sent GMs a memo in mid-June flagging misuse of dugout iPads.
  • Custom apps were feeding real-time recommendations on substitutions, pitch calls, and in-game strategy.
  • The ban took effect this week, at the start of the season's second half.
  • Roughly a third of MLB teams reportedly used the AI-assisted tab at least once.
  • FIFA separately barred AI-assisted calls during live play, mirroring MLB's move.

💬 Smart takes

  • Adam Ottavino, former MLB reliever: said the Mets' AI spending helped push the league toward restrictions.
  • Skeptic: banning the iPad doesn't ban the analytics, teams will just move the same suggestions to a dugout phone or a coach's earpiece.

🧭 Where this goes

  1. Likelyteams shift AI-assisted prep to pre-game only, keeping live decisions human.
  2. Likelyother leagues issue similar in-game AI guardrails within a year.
  3. Possiblea team gets caught circumventing the ban and faces a real penalty.
  4. Wild Carda players' union pushes to formalize AI-assist rules in the next labor deal.

🥄 The Spoon Take

This is the first real AI-in-the-loop rule with enforcement behind it. Sports moves faster than most industries here because a bad call is public in seconds. Every industry debating AI-assisted decisions just got a live test case to watch.

🤔 Pushback

The ban targets one interface, the iPad, the underlying analytics still run everywhere else in the organization.

·Engadget - MLB dugout AI ban·NBC Sports - AI outlawed in dugouts·WBUR - World Cup and MLB AI use
Monday Jul 6
SPORTFIFA

Ball Sensor Costs Croatia A Goal

6JUL
500/SECFIFA

The smart ball made the call, not the ref. FIFA ruled out a Croatia goal against Portugal using ball sensor data. Croatia manager called it game-ruining, but FIFA calls it correct to the millimeter.

Croatia celebrated an equalizer that never counted. The disallowed goal came down to microscopic contact spotted well before impact.

A sensor chip riding inside the Adidas Trionda tracks contact 500 times each second. It flagged Igor Matanovic's touch in the build-up, enough to confirm offside.

Manager Zlatko Dalic called the ruling unfair to players. A separate graphics outage in an earlier match left another offside call unexplained on air.

full brief & sources

⚡ Why this matters

  • The clearest test yet of whether millimeter-precise AI officiating feels fair, even when it's technically right.
  • The ball itself is now a sensor, not just a prop. Officiating moved from the referee's eye to embedded hardware.
  • Sets the tone for how the rest of the tournament handles contested AI calls.

🔍 What happened

  • Croatia vs Portugal, Round of 32: Croatia's apparent equalizer was ruled out for offside.
  • The call used data from a sensor chip inside the Adidas Trionda match ball, sampling 500 times per second.
  • FIFA said the sensor proved contact from Croatia's Igor Matanovic in the build-up, confirming offside.
  • Croatia manager Zlatko Dalic said VAR had ruined the spirit of the game.
  • A separate technical outage in the earlier Qatar-Switzerland match meant the offside graphic never reached broadcast.

💬 Smart takes

  • Zlatko Dalic, Croatia manager: VAR ruined the spirit of the game.
  • FIFA statement: data from the Connected Ball Technology proved contact was made, confirming the offside call was correct.
  • Skeptic: a sensor being technically right doesn't mean fans experience the call as fair. Emotion doesn't read a heartbeat graphic.

🧭 Where this goes

  1. Likelymore marginal offside calls get litigated in public as the tournament goes on.
  2. LikelyFIFA leans harder on ball-sensor data as the tie-breaker in disputed calls.
  3. Possibleplayer and manager pushback grows into a formal complaint about the technology's role.
  4. Wild Carda knockout-stage match gets decided by a call this close, reigniting the whole debate.

🥄 The Spoon Take

The tech is doing exactly what it was built to do: find contact a human eye can't. The problem isn't accuracy, it's trust. A call can be correct to the millimeter and still feel like the game got taken from the players.

🤔 Pushback

FIFA grading its own sensor system as correct has an obvious credibility gap. There's no independent audit of the ball's calibration.

·CNN - Ball sensor explainer·ESPN - FIFA offside ruling·ESPN - Qatar outage report
Monday Jun 29
SPORTOther

Wimbledon Adds An AI Match Assistant

22JUN
FAN ASKSAI ANSWERS

Tennis fans can now just ask what's happening. Wimbledon and IBM added Match Chat, a genAI assistant answering live match questions in plain words. The grass-court major turns AI into a courtside explainer.

Match Chat is the headline feature. Built on IBM watsonx, it sits in the Wimbledon app and site. Ask what happened so far and it answers in conversation, with photos and clips.

Two more tools shipped. Key Moments flags the plays that swung momentum and explains why. Catch Me Up writes AI summaries of each singles player's last and next match.

IBM rebuilt the platform around this. It mapped 15,000 archive assets into a knowledge graph. The AELTC and IBM have partnered for 39 years. Watch if fans treat the bot as the new second screen.

full brief & sources

⚡ Why this matters

  • Live, conversational match explanation moves from broadcast voice-over to an on-demand chatbot.
  • One of sport's oldest events makes genAI a core fan feature, not a gimmick.
  • Sets a template other tournaments and leagues will copy for their own apps.

🔍 What happened

  • Jun 22: Wimbledon and IBM unveiled new AI fan features for the 2026 Championships.
  • Match Chat, built on watsonx, answers live match questions in the app and site.
  • Key Moments identifies momentum-shifting plays and explains the why.
  • Catch Me Up generates summaries for every singles player's recent and upcoming match.
  • IBM mapped 15,000 digital archive assets into a knowledge graph for the platform.

💬 Smart takes

  • IBM: Match Chat lets fans ask what has happened and get instant conversational answers.
  • Skeptic: last year players blasted Wimbledon's AI line-calling, and fan-facing summaries can still get facts wrong.

🧭 Where this goes

  1. Likelyother Grand Slams add conversational match assistants to their apps within a year.
  2. Likelythe bot's summaries become a default second screen during long matches.
  3. Possiblea wrong AI answer during a big match becomes the story instead of the tennis.
  4. Wild Cardlive match chat gets monetized as a premium or sponsored feature.

🥄 The Spoon Take

Sport is becoming the proving ground for everyday AI. Wimbledon isn't replacing the umpire here. It's replacing the friend who explains the match to you. That is a softer, safer place to put genAI than line calls. Watch how fans react when the bot gets one wrong.

🤔 Pushback

Fans may not want to type questions mid-match, and one confident wrong answer would undercut trust faster than the feature builds it.

·IBM - 2026 fan features·AI Magazine - Wimbledon tech hub
Friday Jun 26
SPORTOther

An AI Firm Lands NBA's Richest Patch

25JUN
$50M PATCH

AI compute money is the new sports money. An AI cloud firm just signed the richest jersey patch deal in U.S. team sports, worth over $50M a year.

Two years ago, Iren mostly mined Bitcoin. Now it rents out computing power for AI models. The pivot from crypto to data centers is paying off fast.

The agreement covers the whole franchise, including the WNBA Valkyries and the G League squad. Branding goes up across Chase Center too.

There's a goodwill angle: STEAM education programs and 10,000 youth shirts donated a year. Expect more compute firms chasing arena naming rights next season.

full brief & sources

⚡ Why this matters

  • AI compute cash is now buying the same sports real estate crypto firms chased in 2021.
  • The record number puts a hard dollar value on AI-infrastructure marketing budgets.
  • Signals AI cloud players want consumer brand awareness, not just enterprise sales.

🔍 What happened

  • Jun 25: Iren signed a multiyear patch deal worth over $50M a year.
  • It is the richest jersey sponsorship in North American team sports history.
  • Iren replaces Rakuten as the Golden State Warriors patch partner.
  • The deal spans the WNBA Valkyries and G League Santa Cruz Warriors too.
  • Iren is an Australian AI cloud and data-center company, formerly a Bitcoin miner.
  • Includes STEAM education programs and 10,000 youth jerseys donated annually.

💬 Smart takes

  • Sportico: the deal tops every prior jersey patch in North American team sports.
  • Warriors: framed it around Bay Area AI and STEAM literacy programs.
  • Skeptic: crypto firms bought sports fame too, right before the 2022 crash wiped many out.

🧭 Where this goes

  1. Likelymore AI cloud and chip firms chase NBA patches and arena naming rights.
  2. Likelyjersey patch prices keep climbing as AI money floods in.
  3. Possiblean AI sponsor flames out mid-deal the way crypto sponsors did.
  4. Wild Cardan AI compute firm buys full arena naming rights within a year.

🥄 The Spoon Take

We've seen this movie. In 2021 it was crypto logos on every jersey. Now it's AI compute. The money is real and the deal is huge. But sports sponsorship is where flush industries go to buy legitimacy. The question is whether AI cloud demand lasts longer than crypto's did.

🤔 Pushback

Crypto bought the same jerseys before its crash. A record sponsorship can signal a peak as easily as a durable trend.

·Sportico - Record patch deal·Bleacher Report - Warriors deal·Sportcal - Iren replaces Rakuten
Wednesday Jun 24
SPORTOther

Sportway Raises €20M For AI Cameras

23JUN
AI CAMERA 250K/YR

A Swedish startup is filming the games TV never bothered with. Sportway raised €20M to run 2,500 AI cameras that auto-produce 250,000 youth and lower-league matches a year. No camera crew needed.

Pro-quality coverage used to be reserved for elite leagues. Sportway's AI does the camera work, the editing, and the live feed automatically.

The round values it near €92M, led by Gamma Waves. Founders Jonas Persson and Daniel Franck started it in 2017. It now works with 70+ federations across 21 countries.

The bet: every league wants a broadcast, most can't afford a crew. AI closes that gap and unlocks a long tail of sports media nobody monetized.

full brief & sources

⚡ Why this matters

  • AI moves sports broadcasting from elite-only to every level.
  • The funding signals investors see a real long tail in lower-league media.
  • Camera crews were the cost wall. AI just removed it.

🔍 What happened

  • Jun 23: Sportway raises €20M, led by Gamma Waves.
  • Post-money valuation near €92M, about SEK 1 billion.
  • 2,500 AI camera systems deployed across 21 countries.
  • It produces 250,000+ live events a year.
  • It works with 70+ federations, leagues, and clubs.
  • Founded in 2017 by Jonas Persson and Daniel Franck.

💬 Smart takes

  • Sportway: AI and automation make large-scale broadcasting 'economically viable' from youth leagues to national federations.
  • Skeptic: auto-produced feeds still lag human directors on the big moments. Fine for a U15 game, risky for anything fans pay real money to watch.

🧭 Where this goes

  1. Likelymore federations sign on as the per-event cost keeps falling.
  2. Likelyrivals like Genius Sports push deeper into automated production.
  3. Possiblea major league trials AI cameras as a cheaper second feed.
  4. Wild Carda broadcaster buys Sportway to own the long-tail layer.

🥄 The Spoon Take

Sports media's next growth isn't the Super Bowl. It's the 250,000 games nobody filmed because a crew cost too much. Strip out the humans and the economics flip. Sportway is betting the long tail of sport is bigger than the head.

🤔 Pushback

Auto-produced sport is a commodity. The moment Genius Sports or a hyperscaler bundles it for free, Sportway's €92M valuation looks fragile.

·Tech.eu - Sportway €20M raise·SportBusiness - AI broadcast move·SwedishTechNews - €20M round
Tuesday Jun 23
SPORTOther

One Engineer Rebuilt Wimbledon With AI

22JUN
MONTHS4 WEEKS

A whole sports platform, rebuilt by one person and an AI. For Wimbledon 2026, IBM's build tool pulled 15,000 assets in 47 minutes. Work that took teams months.

IBM calls the tool Bob. It maps and migrates content the way a senior engineer would. One person used it to rebuild the Wimbledon app and website.

The old way needed four or five specialists for months. Bob did the asset extraction in under an hour. New fan tools ride on top: Key Moments explains why a match swung, Match Chat answers questions live.

The story is not tennis. It is how fast a polished product can ship now. One person plus an AI accelerator beat a full team's timeline.

full brief & sources

⚡ Why this matters

  • A real consumer product, rebuilt by one engineer. The build-time math just changed.
  • Sports tech is the showcase. The tool is the news: IBM Bob, an AI dev accelerator.
  • Millions of fans hit this app during the tournament. Not a demo, a live launch.

🔍 What happened

  • Jun 22 - IBM and Wimbledon unveiled new AI fan features for the 2026 Championships.
  • The platform was built with IBM Bob, an internal AI development accelerator.
  • One engineer rebuilt it in 4 weeks. The old way needed 4-5 specialists over months.
  • 15,000 digital assets were extracted and remapped in 47 minutes.
  • New fan tools: Key Moments flags momentum swings, Match Chat answers live questions.
  • Everything runs on watsonx across the Wimbledon app and wimbledon.com.

💬 Smart takes

  • IBM: Bob turned months of specialist work into weeks for one engineer.
  • TechFinitive: the real lesson is build speed, not the fan gimmicks.
  • Skeptic: a clean migration is not a hard product. The wow fades after the first match.

🧭 Where this goes

  1. Likelymore sports and media brands show off AI-built platforms this year.
  2. Likely'one engineer, four weeks' becomes a standard vendor pitch line.
  3. PossibleIBM packages Bob as a paid product, not just an internal tool.
  4. Wild Carda major site relaunch ships in days, not weeks, by end of 2026.

🥄 The Spoon Take

The fan features are nice. The build story is the point. One engineer plus an AI accelerator did what used to need a team and a quarter. Every software timeline you trust just got shorter. The question stops being 'can we build it' and becomes 'why is it not shipped yet.'

🤔 Pushback

Migrating 15,000 assets is mostly plumbing, not product design. The hard, creative parts of software still resist the four-week miracle.

·IBM - Wimbledon 2026 AI features·TechFinitive - The IBM Bob lesson
Monday Jun 22
SPORTOther

Wimbledon Adds AI Match Chat For 2026

22JUN
STARTS JUN 29MATCHAI CHAT

Wimbledon is putting an AI commentator in your pocket. IBM's new Match Chat answers questions during play with stats, photos, and video. A Key Moments tool flags turning points live.

IBM and Wimbledon announced the features for this year's Championships, June 29 to July 12. Match Chat runs on watsonx, trained on Wimbledon's editorial style and tennis language.

Key Moments builds on the Likelihood to Win feature. It scores match momentum from live and historical stats, then surfaces the points that swung the match.

This is fan experience, not officiating. The AI explains the match. It does not call the lines. That is the safe place to deploy sports AI first.

full brief & sources

⚡ Why this matters

  • Sports AI is moving from officiating to storytelling. Lower risk, faster rollout.
  • A conversational AI trained on one tournament's voice is a clean brand play.

🔍 What happened

  • Jun 22 - IBM and Wimbledon unveiled new AI fan features for 2026.
  • Match Chat gives conversational answers with photos and video, built on watsonx Orchestrate.
  • Key Moments scores momentum and flags swing points from live and historical data.
  • Championships run June 29 to July 12. Features live on the app and wimbledon.com.

💬 Smart takes

  • IBM: Match Chat is trained on Wimbledon's editorial style and the language of tennis.
  • Skeptic: AI commentary is fun, but most fans still want the score and a replay.

🧭 Where this goes

  1. Likelyevery major tournament ships a branded AI assistant within two years.
  2. Likelyfan-experience AI scales faster than AI officiating.
  3. PossibleAI commentary becomes a paid premium tier.
  4. Wild Carda league lets an AI generate full live commentary with no human.

🥄 The Spoon Take

Sports AI has two fronts. One calls the game. One explains it. Officiating gets the headlines and the controversy. Fan features like Match Chat get the quiet, steady rollout. Watch the explain side. It ships faster because nobody loses a match over a chatbot.

🤔 Pushback

AI fan tools are easy to launch and easy to ignore, and most viewers still just want the live score and a clean replay.

·PR Newswire - Wimbledon AI features·IBM - Wimbledon 2026
Monday Jun 15
SPORTOther

Rugby Premier League Adds AI Replay

15JUN
CLIPS IN 2 MINAI REVIEWRUGBY

AI replay is spreading past football. India's Rugby Premier League will use Stupa Sports' video review to back referees and doctors from June 16. The selling point is player safety, not just fairer calls.

Stupa Sports runs the replay tech. The Indian firm made its name in table tennis and badminton. Now it's moving into a full-contact sport for the first time.

The system feeds multi-angle camera views to operators. Decisive clips reach officials within two minutes. The biggest use is head injury checks, where speed protects players.

AI officiating started in football and tennis. It's now reaching smaller leagues that want the same credibility. Watch whether the calls hold up in a fast, physical sport.

full brief & sources

⚡ Why this matters

  • AI-assisted officiating is spreading from global leagues to regional ones.
  • The hook here is player safety, not just call accuracy. Head injury checks get faster.
  • Shows AI sports-tech vendors scaling across very different sports.

🔍 What happened

  • Stupa Sports will run its Advanced Decision Review System at Rugby Premier League Season 2.
  • The league plays at Gachibowli Stadium in Hyderabad, June 16 to 28, 2026.
  • Operators feed multi-angle camera views with frame-by-frame replay and zoom to match officials.
  • Decisive clips reach referees and medics within two minutes.
  • Stupa's prior work is in racquet sports: table tennis, badminton, padel, pickleball.

💬 Smart takes

  • Megha Gambhir, Stupa Sports CEO: rugby is high-impact, so decision accuracy and response time matter for fairness and player safety.
  • Sourojit Ghosh, Indian Rugby Football Union: the tech supports foul-play review and head injury assessment protocols.
  • Skeptic: it's operator-led video review, not autonomous AI. The 'AI' label is doing heavy lifting here.

🧭 Where this goes

  1. Likelymore regional leagues add AI-branded review systems to look credible.
  2. Likelyplayer-safety framing becomes the main pitch for officiating tech.
  3. PossibleStupa expands into more contact sports after this rugby debut.
  4. Wild Carda blown review in a high-stakes match sparks a backlash against the tech.

🥄 The Spoon Take

Football and tennis got the AI-officiating headlines. The real spread is happening underneath, in smaller leagues that want the same credibility. The smart move here is leading with player safety, not just fairness. Head injury checks are the easiest version of AI-in-the-loop to defend. Expect that framing everywhere.

🤔 Pushback

This is human operators with good cameras, not autonomous AI. The 'AI-powered' branding may be more marketing than mechanism.

·SportsMint - Stupa ADRS for RPL·Rugby Premier League - Season 2 Hyderabad·Telangana Today - RPL S2 kicks off
Friday Jun 12
SPORTGoogle

Big Tech Wires The World Cup

12JUN
$50B IN BETSAIWORLD CUP

The World Cup kicked off as Big Tech's biggest AI showcase. Lenovo gives all 48 teams an analytics assistant. Google preps 8 squads and fans. Sportradar guards $50B in bets.

The tournament opened June 11 across 16 cities and three countries. Five million will watch in person. Behind them runs the densest computing rollout the sport has seen.

Coaches get a tool that reads hundreds of millions of match data points. Salesforce runs staffing across host cities. Verizon carries the stadium networks.

Integrity software scans for match-fixing across record betting volume. Emergency phone lines get instant translation for visitors. The pitch: automate the measurable, leave judgment to humans.

full brief & sources

⚡ Why this matters

  • The most-watched event on earth is now a live AI stress test.
  • Every vendor wants its AI seen working at planet scale.
  • Shows the split: AI for data and safety, humans for the calls that need judgment.

🔍 What happened

  • Tournament runs June 11 to July 19 across the US, Canada, and Mexico.
  • 48 teams, 104 matches, 16 host cities.
  • Lenovo is FIFA's official tech partner; deal signed October 2024.
  • Google struck team partnerships with 8 squads including the US, Argentina, Brazil, France.
  • Sportradar expects up to $50 billion in betting turnover to monitor.
  • RapidSOS connects 723 million devices for AI-assisted 911 translation.

💬 Smart takes

  • Art Hu, Lenovo CIO: 'Most of the world is watching ... you really have to make sure this works.'
  • Marvin Chow, Google VP: players 'are using AI and digital tools to prepare for matches ... They're people too.'
  • Behshad Behzadi, Sportradar: 'This is the biggest betting event in the world; we expect up to $50 billion overall turnover.'
  • Skeptic: Google admits ticket-hunting agents are 'really in the early days' and won't be ready until the 2027 Women's World Cup.

🧭 Where this goes

  1. LikelyAI broadcast graphics and analytics become standard at every major tournament.
  2. Likelyteams split between trusting the numbers and trusting the coach.
  3. Possiblea visible AI error during a match becomes the story, not the tech.
  4. Possiblebetting-integrity AI flags a real scandal this tournament.
  5. Wild Carda team credits an AI-prepped tactic for a knockout-round upset.

🥄 The Spoon Take

The World Cup became a trade show with a trophy. Every big vendor is paying to show its AI working while 5 billion people watch. The smart move is the restraint. AI does the data and the safety. Humans still make the calls that start arguments.

🤔 Pushback

Most of this is sponsorship theater. The AI that matters runs in the background, and fans will remember the goals, not the analytics tools.

·Fortune - Big Tech's World Cup·PYMNTS - AI from first whistle·TechTimes - Tech's biggest live test
Wednesday Jun 10
SPORTOther

ESPN Pulls Its AI Portraits Mid-Finals

10JUN
GONE IN 4 DAYSAI PORTRAITFANS

ESPN aired an AI-animated Tony Parker during the NBA Finals. Fans hated the warped face, and the portraits are gone for the rest of the series. Backlash killed the feature in four days.

The animation ran before an ad break - a 2003 championship photo brought to life with a smile the original never had. Bill Russell and Kobe Bryant got the same treatment that night.

The network first told Front Office Sports it was 'evaluating' the technology. Two nights later it was dead, with no plans to bring it back.

Sports audiences are the sharpest AI-slop detectors anywhere. Caitlin Clark called out her own team's AI graphics last month. Shipping fake-looking content now costs a brand within hours.

full brief & sources

⚡ Why this matters

  • A major broadcaster reversed an AI feature in four days, on its biggest stage of the year.
  • The backlash was not about jobs or ethics - the output just looked wrong to people who know these faces.
  • Every media brand testing generative content on air now has a public cautionary tale.

🔍 What happened

  • Game 1, June 4: ESPN aired an AI-animated portrait of Spurs legend Tony Parker before an ad break.
  • Parker's face did not match the original 2003 championship photo - fans spotted it instantly and the clip went viral.
  • Two more AI portraits aired the same night: Bill Russell in the 1960 Finals and Kobe Bryant in 2001.
  • ESPN confirmed to Front Office Sports the images were made with AI tools and said it was 'evaluating' their use.
  • By Game 2 the portraits were gone. A source confirmed they will not return for the remainder of the series.

💬 Smart takes

  • Jon Healy, sports journalist: 'Could ESPN really not find a genuine shot of Tony Parker as they cut to an ad break? Just had to use AI.'
  • Colin Salao, Front Office Sports: the portraits are 'just the latest illustration of the prevalence of AI in the world of sports - and fans' backlash to it.'
  • Skeptic: this is a graphics-package misstep, not an AI verdict - ESPN keeps using AI across production tools nobody complains about.

🧭 Where this goes

  1. Likelynetworks keep AI in back-office production and away from on-air talent likenesses.
  2. Likelyleagues add likeness-approval language for AI-altered imagery in upcoming media rights deals.
  3. Possiblea broadcaster ships a clearly disclosed, opt-in AI segment that fans accept within a year.
  4. Wild Carda player sues a network over an AI-altered likeness aired without consent.

🥄 The Spoon Take

Sports fans just ran the fastest AI QA cycle on record. ESPN shipped, the internet rejected, the feature died in four days. The lesson is not 'don't use AI' - it's that audiences punish output that disrespects faces they know by heart.

🤔 Pushback

ESPN quietly keeps AI everywhere else in production - one pulled graphics package says little about where broadcast AI actually lands.

·Front Office Sports - AI graphics ended·Front Office Sports - Parker backlash·Yahoo Sports - Portraits pulled
Tuesday Jun 9
SPORTFIFA

AI Calls Offside At The World Cup

7JUN
FIRST TIMEAIREFEREE

AI will call every offside at the 2026 World Cup. The system tracks each player with 29 cameras and signals the referee in seconds. It's the first time AI is the final word on a call in a major tournament.

The system is called SAOT — Semi-Automated Offside Technology. It watches each leg, arm, and shoulder, and pings the assistant referee's earpiece when a player is more than 10cm beyond the defender.

There's no replay review on top, so the AI's call is final. Lenovo built the camera system, FIFA owns the deployment. Football's most-argued judgment just became the cleanest live test of AI calling the shots.

Worth watching how fans react when AI disallows a goal the eye missed. The whole 'AI in the loop' rollout in other sports follows from this test.

full brief & sources

⚡ Why this matters

  • SAOT is the first AI system deployed as the sole arbiter — not advisory — of an offside call at a World Cup.
  • The decision to remove human VAR override is the meaningful policy shift, not the technology itself.

🔍 What happened

  • FIFA's Semi-Automated Offside Technology reads 29 dedicated limb-tracking cameras per stadium at 50fps.
  • The model maps 29 body points per player, freezes attacker position when the ball is played, computes the 3D offside line, and routes the call to the referee's earpiece.
  • Full pipeline runs in under 3 seconds. Rejected goals show a 3D reconstruction on stadium screens within 20 seconds.

💬 Smart takes

  • The earpiece as output interface is worth studying: the AI bypasses screen-based UI entirely and speaks directly to the decision-maker.
  • The 10cm precision number is a product communication lesson — one concrete stat that shifted public trust faster than any explainability diagram could.
  • The 3D replay is the explainability layer: instead of logs or confidence scores, the system shows the stadium what it saw.

🧭 Where this goes

  1. World Cup 2026, co-hosted by USA, Canada, Mexico.
  2. SAOT infrastructure installed across all 16 host stadiums.
  3. First live use at scale — previously piloted at 2022 Qatar World Cup in limited form.

🥄 The Spoon Take

What's interesting: the referee still raises the flag. But the judgment moved. This is the cleanest 'is the call objective enough for AI' test high-stakes industries have run. How FIFA handles the first controversy will shape every AI-in-the-loop rollout that follows.

🤔 Pushback

The first goal wrongly disallowed becomes a cultural moment — the tech can be right 99.9% of the time, and the 0.1% will define the narrative.

·FIFA SAOT Technical Brief·The Athletic — How SAOT Changed Officiating at World Cup 2026