Public Knowledge’s Post

Is OpenAI crossing the line by mimicking celebrities without consent? Can Altman still be trusted? ➡️ Scarlett Johansson was "shocked, angered, and in disbelief" when she discovered OpenAI's ChatGPT 4o used a voice eerily similar to hers for its new system, "Sky." ➡️ Despite declining their offer to use her voice, Johansson found herself battling the unauthorized mimicry. OpenAI insists a different actress was used, but the voice’s uncanny resemblance to Johansson's in the film Her led to public outcry and legal action. ➡️ Johansson met with OpenAI in September, where CEO Sam Altman pitched the idea of using her voice to bridge the gap between tech and creatives. ➡️ She declined, yet weeks later, Sky's voice left her friends and the media unable to distinguish between her real voice and the AI-generated one. Altman's cryptic "Her" tweet only fueled suspicions of intentional mimicry. ❓ As the legal dust settles, it begs the question: Can Sam Altman be trusted? Read the full story on RollingStone: https://lnkd.in/gwAMVMd3 #Trust #OpenAI #Altman #Ethics #ResponsibleAI ---- 💡 𝗜𝗳 𝘆𝗼𝘂 𝗲𝗻𝗷𝗼𝘆𝗲𝗱 𝘁𝗵𝗶𝘀 𝗰𝗼𝗻𝘁𝗲𝗻𝘁, 𝗯𝗲 𝘀𝘂𝗿𝗲 𝘁𝗼 𝗱𝗼𝘄𝗻𝗹𝗼𝗮𝗱 𝗺𝘆 𝗻𝗲𝘄 𝗮𝗽𝗽 𝗳𝗼𝗿 𝗮 𝘂𝗻𝗶𝗾𝘂𝗲 𝗲𝘅𝗽𝗲𝗿𝗶𝗲𝗻𝗰𝗲 𝗯𝗲𝘆𝗼𝗻𝗱 𝘆𝗼𝘂𝗿 𝘁𝗿𝗮𝗱𝗶𝘁𝗶𝗼𝗻𝗮𝗹 𝗻𝗲𝘄𝘀𝗹𝗲𝘁𝘁𝗲𝗿 - you can have real-time insights, recommendations (a lot more than I share here) and conversations with my digital twin via text, audio or video in 28 languages! Join >5000 users who went before and go to app.thedigitalspeaker.com to sign up and take our connection to the next level! 🚀

From Reuters: Johansson's accusation that the ChatGPT-maker copied her performance in the Spike Jonze-directed feature film, after failing to strike an agreement, rekindled the creative class’s anxiety about the existential threat posed by #AI, even as Hollywood studios test new tools and mull alliances with OpenAI. “This seemed to strike a real chord,” said one industry executive. “It kind of puts a human face on it … There’s a well-known tech company that did something to a person we know.” OpenAI stunned the world in February with feature film-like quality videos generated by its text-to-video tool, Sora. Since then, Hollywood executives and agents have met the company multiple times to discuss potential creative partnerships and applications of the technology, according to agents and industry executives. OpenAI CEO Sam Altman said in a statement Monday that the voice "is not Scarlett Johansson's, and it was never intended to resemble hers. We cast the voice actor behind Sky’s voice before any outreach to Ms. Johansson." The company, whose largest investor is Microsoft did not reply to requests for comment on its relationship with #Hollywood after the dispute. Even before the latest conflict, agents and executives who spoke with Reuters on condition of anonymity have said for weeks they are concerned that OpenAI’s models appear to have been trained on copyrighted works, which the #tech company deemed as a fair use because they are publicly available on the internet. That is seen as a major obstacle by some professional directors and filmmakers, who may be reluctant to use a tool built, without consent, on others’ work. But technologists in the #entertainment industry view Sora as a promising potential tool to augment the #film- and TV-making process. They see near-term applications for the technology to accelerate the pace of digital effects. Fox Corporation already uses OpenAI’s ChatGPT to recommend new #TV shows and movies for viewers of its Tubi #streaming service. Although OpenAI has said it aims to protect copyrights - blocking the ability to generate videos featuring known characters like Superman or prominent actors like Jennifer Aniston -- there remain concerns about how it will safeguard lesser-known performers.

"Scarlett Johansson Takes the AI Fight to Big Tech, and Big Media Should Follow" - read my latest column in TheWrap. Here's an excerpt: Does anyone really believe OpenAI CEO Sam Altman anymore?  Altman launched #OpenAI in 2015 with a mission to develop AI that “benefits all of humanity.” Fast forward nine years later, humanity be damned — Altman and his ego needed a “Her” moment when OpenAI launched its latest version, ChatGPT-4o. Problem is, Scarlett Johansson was having none of it. Neither is much of the creative community, which is finally beginning to understand its collective power and is rising up to say, “Stop the madness!” That “madness” is the arrogant taking of not just creative works — which is the subject of an ever-growing number of infringement cases in the courts – but now also the taking of creators’ personas themselves. Exposed were Altman’s true colors when he asked Johansson for permission to use her voice, she declined, then he asked again just two days before OpenAI’s big ChatGPT-4o launch, and went forward anyway when he didn’t hear back.  Altman and his apologists want us to believe that there is no harm, no foul here because OpenAI reportedly used a faux Johansson voice, not her actual one. But that doesn’t diminish Altman’s theft, both ethically and legally. We know what Altman was really doing, and a wide body of established NIL (name, image and likeness) law protects celebrities from companies using “sound-alikes” for commercial purposes. Altman’s humiliating gaffe at the hands of Johansson may end up being a watershed moment in Big Tech’s relentless quest to add the next trillions of dollars to its market capitalizations. Her public pushback shines Hollywood’s bright spotlight on what’s really going on here – and it certainly ain’t Big Tech concerns for trust and safety. If there were any doubt about that, the recent departures of OpenAI’s two top executives on that front – and the dissolution of their entire “risk mitigation” team — erased it. Not a good couple weeks for Sam. Now it’s time for “Big Media” – the major studios, streamers, record labels, game companies, publishers – to follow Johansson’s lead and magnify the issue of Big Tech’s wholesale theft. Unified collective action can best lead to both a more equitable sharing of the generative AI pie going forward and significant payments for Big Tech’s past misdeeds of scraping copyrighted works without consent or compensation. And creators, one more thing. The licensing deals Big Tech is offering today – like OpenAI’s reported $250 million five-year licensing deal with News Corp – may sound like big numbers today. But my bet is that those numbers will seem like trifles to the overall value generated by Big Tech with your IP over that period of time. https://lnkd.in/gzKakk9Y #media #entertainment #hollywood #AI #generativeAI AI LA #music #writers The New York Times

You know how GenAI can create problems even faster than humans can? Here's one. ;-) Spoiler alert. This report concludes (emphasis is mine): "From my testing across 10 publications, it appears that currently, ChatGPT is often *doing what its predictive text generation does best*: predicting the most likely version of the URL for a given story — rather than the correct one." Val Swisher Here's an example where the content has already been curated and cleaned up (props to the journalists!), just like Content Rules, Inc. recommends for any organization that wants to get useful results from an AI. The AI can still get it wrong because it's very good at finding new ways to do things that humans do, only faster and more visibly. Gotta keep an eye on the AI in these early days!

"#OpenAI and #TimeMagazine strike multiyear deal to improve ChatGPT with journalistic content." 🗞️ While the #lawsuit between OpenAI and The New York Times (#NYT) is still ongoing and we know some more about: ✔️ The New York Times has spent approximately $1 million so far in its lawsuit against OpenAI and Microsoft but the newspaper has stated that these expenses do not impact its operational performance; ✔️ NYT alleges that its articles are generally licensed at $10 each and that the dataset includes at least 16 million unique records, potentially $160M in damages. If the infringement is deemed willful, statutory damages could be up to $150,000 per copyrighted work; ✔️ OpenAI / Microsoft have both filed motions to dismiss parts of the lawsuit; OpenAI "announced a "multi-year content deal" that will allow OpenAI to access current and archived articles from more than 100 years of Time's history." 📌 As part of the deal, as reported by CNBC (https://shorturl.at/nFoG0): ✔️ The Microsoft-backed startup will be able to display Time's content in its ChatGPT chatbot and use Time's content "to enhance its products," or, likely, to train its artificial intelligence models; ✔️ A similar partnership announced by OpenAI and #NewsCorp. in May allows OpenAI to access current and archived articles from News Corp.'s outlets, including The Wall Street Journal (#WSJ), #MarketWatch, #Barron's, The New York Post (#NYP) and more. There are also the following as recalled by CNBC: ✔️ In 2023, a group of prominent U.S. authors, including JF, JG, George R.R. Martin and JP, sued OpenAI alleging © infringement in using their work to train ChatGPT (https://shorturl.at/cZVtX) ✔️ In July, two authors filed a similar lawsuit against OpenAI, alleging that their books were used to train the company's chatbot without their consent (https://shorturl.at/8R2Nn). 💡A licensing agreement is a good strategy for news organizations, but it is the only one? And for you? In a short list between lights and shadows: 🟩Financial compensation: A licensing agreement ensures that news organizations receive financial compensation for the use of their content by AI technologies; 🟩Copyright protection: News organizations can establish clear terms for the use of their content, thereby protecting their IP rights; 🟩Collaboration, Innovation: It can facilitate collaboration between news organizations and tech companies, promoting innovation and the development of new products and services that can benefit both parties. 🟥While a licensing agreement can be advantageous, it is not the only solution: news organizations can also choose to pursue legal action to protect their rights; 🟥Legal actions can be costly and time-consuming but they can also lead to significant compensations and establish legal precedents; 🟥 Risk of legal precedents: If a legal case is lost, it could set a precedent that makes it more difficult for news organizations to protect their content in the future.

🔮 Prediction: If Scarlett Johansson decides to pursue legal action against OpenAI for allegedly creating a ChatGPT voice that closely resembles hers, the case could potentially spark a significant debate within the legal and tech communities. - The claim of breaching her right to publicity raises questions about the boundaries of AI technology and its implications for personal likeness rights. - It could set a precedent for future cases involving the intersection of celebrity identities and artificial intelligence. - The outcome of this potential legal battle may have far-reaching consequences for how AI developers approach creating virtual entities that resemble real individuals.

Renowned authors, with support from the Author's Guild, allege OpenAI is making them "unwilling accomplices in their own replacement." According to the class-action lawsuit, defendants are seeking redress for “flagrant and harmful infringement” on their registered copyrights. “These algorithms are at the heart of [OpenAI's] massive commercial enterprise. And at the heart of these algorithms is *systemic theft on a mass scale*.” Axios #copyrightlaw #ChatGPT #AI https://lnkd.in/efFfHS4P

Microsoft and OpenAI face lawsuits from multiple news organizations. #TechLawsuits 🤝 Follow us on Discord 🔜: https://lnkd.in/gt823Zd3 🤝 Follow us on Whatsapp 🔜 https://wapia.in/wabeta _ ❇️ Summary: Microsoft and OpenAI are facing lawsuits from news organizations such as The New York Times, Raw Story, The Intercept, and AlterNet for the use of ChatGPT and Copilot, claiming that their stories are being copied without proper attribution. OpenAI has filed a counterclaim against The New York Times, alleging that the newspaper hired someone to hack into their services. These legal battles could determine the rights of news organizations to protect their content from generative AI services. Hashtags: #chatGPT #MicrosoftLawsuits #OpenAILawsuits

Do you think the voice of Scarlett Johansson is part of OpenAI’s training dataset? 🤔 The actress is experiencing the unsettling reality of the film “Her” without her consent. Despite explicitly asking OpenAI not to use her voice, the new ChatGPT voice, Sky, sounds strikingly similar to hers. This raises significant concerns about the ethical use of AI and the protection of creators’ rights. 👉 At Fair AI Data, we are committed to transparency and fairness. Creators deserve to have their voices and work respected and protected. It’s alarming to see AI companies continually finding new ways to use user data without consent, extending beyond just data used for training. It’s time to set a new industry standard where companies train on ethically-sourced data and creators get accredited and compensated. Because the copyright infringement cases & claims keep piling up ✊ Read more about Scarlett Johansson’s case: https://lnkd.in/dYbycm4P

ChatGPT to summarize Politico and Business Insider articles in ‘first of its kind’ deal OpenAI to pay German media group Axel Springer to use its material, including stories behind paywalls. https://lnkd.in/enbFU5K2 #OpenAI #ChatGPT #AxelSpringer #MediaInnovation #AIJournalism #BusinessInsider #Politico #TechNews #DigitalMedia #AIinMedia"