The recent trend of OpenAI partnering with major media companies has raised some curiosity and concern. OpenAI, the maker of ChatGPT and other large language models (LLMs), has forged alliances with seven major media outlets in less than a year, including well-known publishers like News Corp. and German publisher Axel Springer. While the exact terms of these deals have not been disclosed, it is estimated that OpenAI is paying tens of millions of dollars for access to the media produced by these publishers.
One of the main reasons for these partnerships is to gain access to licensed training data that can be used to improve AI models like ChatGPT. By partnering with media companies, OpenAI can build powerful new AI models that can generate content as well as a professional journalist. Additionally, OpenAI could potentially train its generative AI video model, Sora, to create documentary-style content from text prompts.
However, the decision to pay for licensed content instead of scraping publicly available data for free is likely due to the pushback from artists, creatives, and media companies who view scraping as ethically questionable. OpenAI has introduced a code that website owners can use to prevent their sites from being scraped and trained on by AI models. Furthermore, OpenAI plans to create a Media Manager product that allows content creators to flag their work and prevent it from being ingested by AI scrapers. These measures aim to address concerns about copyright infringement and ethics.
For the media companies involved, one benefit of partnering with OpenAI is increased exposure. ChatGPT will surface their articles among its outputs, providing more visibility to their content. However, it remains unclear how much referral traffic ChatGPT is driving to these publications or how much it will directly quote from articles instead of summarizing them. This raises concerns about the potential loss of ad revenue and paying subscribers for media outlets if users can access the information they need through ChatGPT.
This partnership between OpenAI and media companies is reminiscent of the changes brought about by tech giants like Google, Facebook, and Twitter. These platforms became major sources of referral traffic for publishers, leading to shifts in audience and revenue. Now, with the rise of generative AI, media companies are once again seeking partnerships with tech firms to adapt to changing media habits. However, the history of tech companies changing strategies and tools unpredictably raises questions about the long-term sustainability of these partnerships.
While OpenAI may benefit from gaining access to more factual information, it is uncertain whether users will rely on ChatGPT as a news source or aggregator. The decline in news consumption rates in the US and the rise of misinformation on digital platforms highlight the challenges faced by traditional media outlets. OpenAI’s exploitation of the lack of direct funding for media allows it to gain more power and information. Whether ChatGPT becomes the “homepage of the internet” or if users will prefer other sources for news remains to be seen. Ultimately, the outcome of these partnerships depends on whether media companies are getting their money’s worth and if users find value in accessing news through AI models.