Apple is set to unveil its Private AI Cloud, named Private Cloud Compute, next week, marking a significant step in the company’s commitment to enhancing security and user privacy in artificial intelligence applications.
As part of this initiative, Apple is implementing a bounty programme designed to incentivize security researchers to identify vulnerabilities within its private AI cloud infrastructure.
The bounty programme offers financial rewards for uncovering critical security flaws.
Researchers can earn up to $1 million for reporting exploits that enable the remote execution of malicious code. Additionally, the program provides up to $250,000 for exploits that could extract sensitive user information or customer prompts, and up to $150,000 for vulnerabilities that access sensitive data from a privileged network position.
Apple emphasized its focus on addressing significant security issues that could impact user data and inference requests beyond the trust boundary of the private cloud.
This initiative is an extension of Apple’s ongoing bug bounty programme, which aims to encourage researchers to report security flaws and bolster the overall security of its products.
Previously, Apple introduced a special iPhone exclusively for researchers, designed to enhance defenses against spyware and improve overall security measures.
In an effort to promote transparency and build trust, Apple has released extensive documentation detailing the security features of its Private Cloud Compute service, including source code.
The company reiterated its commitment to user privacy, asserting that the new cloud service will efficiently handle demanding AI tasks while safeguarding personal data through its Apple Intelligence model.
