According to a press release issued on Friday, Apple has signed the White House's voluntary commitment to develop safe, reliable, and trustworthy artificial intelligence. This move signifies that Apple will soon integrate its generative AI product, Apple Intelligence, into its core products, serving Apple's 2 billion users.
In July 2023, Apple joined forces with 15 other tech companies, including Amazon, Google, Meta, and Microsoft, to pledge adherence to the White House's basic rules for generative AI. Although Apple did not disclose specific plans to integrate AI technology into iOS at the time, the company clearly stated at the June Worldwide Developers Conference (WWDC) that it would fully commit to the field of generative AI, starting with partnering to embed ChatGPT in iPhones.
Apple's voluntary commitment to the White House, while limited in binding force, marks its first step in the AI domain. The White House stated that this is the "first step" for Apple and other companies to develop safe, reliable, and trustworthy AI. Following this will be President Biden's AI executive order in October, with several bills currently under review in federal and state legislatures to better regulate AI models.
Under the commitment, AI companies will conduct red team testing on AI models before public release and share this information with the public. Additionally, companies must keep the weights of unreleased AI models confidential and study these weights in a secure environment, limiting access to the model weights. Finally, companies agree to develop content labeling systems, such as watermarks, to help users distinguish between AI-generated content and non-AI-generated content.
The U.S. Department of Commerce has indicated that it will soon release a report on the potential benefits, risks, and impacts of open-source foundational models. Open-source AI is becoming a politically charged regulatory battleground. Some factions want to restrict access to the model weights of powerful AI models for safety reasons, which could limit the development of the AI startup and research ecosystem. The White House's stance in this area could have a significant impact on the entire AI industry.
The White House also noted that federal agencies have made significant progress on the tasks outlined in the October executive order. To date, federal agencies have hired over 200 AI-related personnel, granted access to computing resources to more than 80 research teams, and released multiple frameworks for developing AI.