(San Francisco) Apple has joined a dozen tech companies that have pledged to follow certain rules to limit the risks of so-called generative artificial intelligence, the White House said Friday.
In early June, Apple unveiled Apple Intelligence, a new nomenclature that should optimize the use of its devices, from the iPhone to the Mac, thanks to generative AI.
The Apple brand’s announcement also included a partnership with OpenAI and the availability of ChatGPT for iPhone users.
Generative AI, which can quickly produce images, sounds or even videos on simple request in everyday language, gives rise to hopes of great progress, particularly in medicine, but also fears of an explosion of disinformation, massive job losses or even the theft of intellectual property. Not to mention the use that authoritarian regimes or criminal organizations can make of AI.
Amazon, Google, Microsoft and OpenAI had already joined the call launched by the American authorities a year ago to self-regulate.
President Joe Biden’s administration said at the time that it had secured commitments from the companies “to help ensure the safe, secure and transparent development of AI technology.”
Safeguards include attack simulations of artificial intelligence models to detect vulnerabilities.
According to the White House, these tests should focus on societal risks and national security issues, such as cyberattacks or the development of biological weapons.
The companies that signed have committed to sharing best practices with each other and with legislators, researchers and associations to make these new systems less dangerous.
In the spring, a new federal council was formed to advise the U.S. government on the “safe and secure” use of artificial intelligence (AI). It includes the heads of OpenAI, Microsoft, and Google.