How NBC Plans To Use AI To Capture Nostalgia For Its NBA Broadcast

Former NBA basketball player Magic Johnson, NBC commentator, at the 1993 NBA Finals – Game 5, ... More Chicago Bulls vs Phoenix Suns, held at the Chicago Stadium, Chicago, Illinois, US, 18th June 1993. (Photo by Bongarts/Getty Images) AI is breathing new life into one of sports media’s most iconic partnerships: the NBA and NBC. Last summer, the two sides inked a new 11-year deal that will bring the NBA back to NBCUniversal networks like NBC, Telemundo, and streaming service Peacock. Now, NBC is adding an emotional and technological twist to the revival. NBC revealed that the voice of Jim Fagan will be used to during promotions and broadcasts courtesy of AI technology. Fagan, whose deep baritone became synonymous with Michael Jordan-era broadcasts, will once again narrate promos, title sequences, and select programming, despite his passing in 2017. Jim Fagan’s Legacy And The 1990s NBA Jim Fagan was synonymous with the Michael Jordan era of the NBA. His resonant voice served as the gateway to countless iconic moments on NBC’s broadcasts from 1990 to 2002. As the narrator behind the unforgettable “This is the NBA on NBC” intro, Fagan’s tone set the emotional stage for epic rivalries, Finals showdowns, and the rise of global superstars like Shaquille O’Neal, Hakeem Olajuwon, and Kobe Bryant. His voice became as much a part of the ’90s NBA experience as the soaring dunks and physical defense. NBC released a teaser video earlier today with a preview of the AI-generated voice of Jim Fagan accompanying a sizzle reel of flashy in-game highlights. NBC also reacquired the rights to the “Roundball Rock” theme written by John Tesh. With its return alongside Jim Fagan’s AI-recreated voice, NBC is doing more than just reviving old assets; it’s resurrecting a cultural moment that defined a generation of basketball viewing. How AI Is Used For Voice Generation In Sports This is not NBC’s first foray into AI-generated voiceovers, and the Fagan announcement is part of a broader play to integrate AI into story telling. During the 2024 Summer Olympics, fans could watch a personalized daily highlights narrated by Al Michaels. NBC’s use of an AI-generated version of Al Michaels’ voice received a notably positive reception. The response was largely positive. Highlighted in a review by The Washington Post, the AI voice of Al Michaels was described as “surprisingly good,” noting that while the synthesized voice sounded slightly flatter than the original, it effectively captured Michaels’ distinctive cadence and delivery. The AI even managed to pronounce complex athlete names correctly, enhancing the viewing experience for fans unfamiliar with certain sports. PITTSBURGH, PA – NOVEMBER 02: Al Michaels looks on from the sideline prior to an NFL football game ... More between the Tennessee Titans and the Pittsburgh Steelers at Acrisure Stadium on November 2, 2023 in Pittsburgh, Pennsylvania. (Photo by Cooper Neill/Getty Images) How AI Generates Voices AI-generated voices rely on a branch of artificial intelligence called neural text-to-speech (or neural TTS), which uses deep learning to synthesize human-like speech from text. At the heart of this process is a machine learning model trained on hundreds or thousands of hours of real voice recordings. These recordings teach the AI not only how a person sounds, but how they pace, emote, emphasize, and even breathe during speech. Convert text to speech concept. Translation of a text message into a voice audio message using ... More neural networks. Text-to-Speech vector illustration for app, service, landing pages. Vector illustration. To recreate the voice of a specific individual, companies typically use a technique called voice cloning. This involves feeding the AI model a rich dataset of archival audio from that person, allowing it to learn vocal nuances and inflections. The model then generates speech from new text inputs that mimics the speaker’s tone and style. Advanced systems like those used by NBC often include phoneme-level alignment to capture individual sounds of speech and prosody modeling to mimic human pitch, loudness, tempo, and rhythm. This makes the output feel less robotic and more human. Oklahoma City Thunder’s Shai Gilgeous-Alexander (2) works for a shot attempt as Denver Nuggets’ ... More Jamal Murray, left, defends in the second half of Game 1 of an NBA basketball second-round playoff series Monday, May 5, 2025, in Oklahoma City. (AP Photo/Nate Billings) It is likely that human oversight plays a crucial role. Editorial teams vet scripts, ensure accurate pronunciations (especially with international athlete names), and fine-tune intonation. Takeaway The NBA has been making strides in AI and analytics over the last few years, and as NBC prepares to relaunch its NBA coverage, the use of AI-generated voices like Jim Fagan’s is not just a technological gimmick. By combining the emotional resonance of familiar sounds with the flexibility of AI, networks are finding new ways to deepen fan engagement and scale storytelling like never before. If the return of “Roundball Rock” and Fagan’s voice signals anything, it is that nostalgia, when powered by innovation, can be a foundation for the next era of sports broadcasting. Source: https://www.forbes.com/sites/giovannimalloy/2025/05/06/how-nbc-plans-to-use-ai-to-capture-nostalgia-for-its-nba-broadcast/