OpenAI bribing publishers?

According to a leaked slide deck, OpenAI has a “Preferred Publishers Program” (PPP) that they pitch to news publishers, like the Financial Times and Le Monde, so they can use their content for AI model training, without getting sued for copyright infringement, like they have been by some news outlets, like the New York Times, for example.

What is the Preferred Publisher Program?

The slides reveal that OpenAI’s PPP is just for “select, high-quality editorial partners” as their main aim with these partnerships is to help ChatGPT users get access to and engage with, only the very highest-quality content.

What do publishers get from the Preferred Publisher Program?

OpenAI is incentivizing publishers to join the program by promising them priority placements, “richer brand expression”, and more prominent links in ChatGPT responses.

Publishers will also get two forms of financial compensation:

1. a guaranteed payment, which is effectively a licensing fee for allowing OpenAI to access its backlog of content

2. a variable payment, which is based on how many users interact with their linked or displayed content.

What does this mean for the content we consume?

This type of incentivized partnership, which places certain content above other content, has sparked concern that the information we get shown by AI models, like ChatGPT, may not be the most relevant or useful, but could be biased based on marketing spend.

‍