OpenAI says it's building a tool to let content creators 'opt out' of AI training | TechCrunch

🌈 Abstract

The article discusses OpenAI's development of a tool called "Media Manager" to allow creators and content owners to better control how their content is used in training generative AI models. The goal is to have this tool in place by 2025, as OpenAI works with creators, content owners, and regulators to establish a standard for this process.

🙋 Q&A

[01] OpenAI's Media Manager Tool

1. What is the purpose of OpenAI's Media Manager tool?

• The tool will allow creators and content owners to identify their works to OpenAI and specify how they want those works to be included or excluded from AI research and training.
• The goal is to have this tool in place by 2025 as OpenAI works with "creators, content owners and regulators" toward a standard.

2. What are the key features of the Media Manager tool?

• It will use "cutting-edge machine learning research" to identify copyrighted text, images, audio, and video across multiple sources and reflect creator preferences.
• Over time, OpenAI plans to introduce additional choices and features for creators and content owners.

3. What is the motivation behind developing the Media Manager tool?

• It is a response to growing criticism of OpenAI's approach to developing AI, which relies heavily on scraping publicly available data from the web.
• This includes recent lawsuits from U.S. newspapers accusing OpenAI of using their articles to train generative AI models without compensation or credit.

4. What are some of the current limitations of OpenAI's approach to handling creator content?

• The current opt-out workflow for images is described as onerous, requiring creators to submit each image individually.
• OpenAI reportedly pays relatively little to license content.
• OpenAI's current solutions don't address scenarios where creators' works are quoted, remixed or reposted on platforms they don't control.

[02] Broader Landscape of Tools for Generative AI Content Control

1. What are some other tools and initiatives aimed at giving creators more control over their content in generative AI?

• Startup Spawning AI offers an app to identify and block bot scraping attempts, as well as a database where artists can register their works to disallow training by vendors.
• Steg.AI and Imatag help creators establish ownership of their images by applying watermarks.
• Nightshade, a project from the University of Chicago, "poisons" image data to make it unusable for AI model training.

2. What are the key challenges these tools and initiatives are trying to address?

• Providing creators with more control and options for how their content is used in training generative AI models.
• Establishing ownership and provenance of creative works to prevent unauthorized use in AI training.
• Developing technical solutions to block or disrupt the scraping of content for AI model training.