Insights from President Biden's Executive Order: The Imperative Role of Confidential Computing in Trustworthy AI Development

The recent Executive Order issued by President Biden underscores the criticality of preserving privacy, data integrity, and security to enable safe and responsible AI development and use. As AI continues to grow and evolve, the need for robust protective measures has gained prominence. The order specifically mentions: 

“(f)  Americans’ privacy and civil liberties must be protected as AI continues advancing.  Artificial Intelligence is making it easier to extract, re-identify, link, infer, and act on sensitive information about people’s identities, locations, habits, and desires.  Artificial Intelligence’s capabilities in these areas can increase the risk that personal data could be exploited and exposed.  To combat this risk, the Federal Government will ensure that the collection, use, and retention of data is lawful, is secure, and mitigates privacy and confidentiality risks.  Agencies shall use available policy and technical tools, including privacy-enhancing technologies (PETs) where appropriate, to protect privacy and to combat the broader legal and societal risks — including the chilling of First Amendment rights — that result from the improper collection and use of people’s data.”

This is where Confidential Computing and Secure Enclaves can play a critical role, offering a viable solution to ensure trust and security that is indispensable for fostering a secure AI landscape. The Executive Order specifically mentions these technologies in the following section:

“(z) The term “privacy-enhancing technology” means any software or hardware solution, technical process, technique, or other technological means of mitigating privacy risks arising from data processing, including by enhancing predictability, manageability, disassociability, storage, security, and confidentiality. These technological means may include secure multiparty computation, homomorphic encryption, zero-knowledge proofs, federated learning, secure enclaves, differential privacy, and synthetic-data-generation tools. This is also sometimes referred to as “privacy-preserving technology.”

What does the new Executive Order mean to agencies and enterprises?

[.color-green]Protect individuals' privacy[.color-green]

As mentioned before, the Order first and foremost stresses the significance of protecting individuals' privacy amidst the proliferation of AI. To do so, it advocates for the use of privacy-enhancing technologies (PETs). Confidential Computing, as a prime example of such technologies, aligns with this goal by preserving the confidentiality of data during processing. This quality makes it possible to train AI systems on encrypted data without compromising the privacy of the underlying information, thereby playing a central role in safeguarding individuals' privacy rights.

[.color-green]Advance equity and civil rights[.color-green]

The Order also emphasizes the need to address issues related to equity and civil rights, advocating for the prevention of algorithmic discrimination. For agencies to reach this objective, they must find ways to ensure data and code integrity when developing AI systems. In that context, Confidential Computing can contribute to the development of these AI systems that are unbiased and free from discriminatory practices, in no small measure thanks to its unique remote attestation capabilities that offer code integrity guarantees.

[.color-green]Promote innovation and competition[.color-green]

Regarding the Order's focus on promoting innovation and competition, Confidential Computing serves as a catalyst for driving AI research and development. By facilitating the secure sharing of data and models, Confidential Computing fosters an environment where AI innovation can flourish without compromising privacy. This is crucial for encouraging small developers and entrepreneurs to participate in the AI ecosystem, thereby promoting a fair and competitive landscape for AI advancements.

[.color-green]Share safety test results with the U.S. government[.color-green]

Finally, one of the directives of the Executive Order focuses on establishing new standards for AI safety and security. It emphasizes the need for developers of advanced AI systems to share crucial safety test results with the government, reinforcing the importance of comprehensive and transparent safety measures. In this context, Confidential Computing provides a straightforward path to share data safely and freely with the government when necessary, thus mitigating potential risks to national security, and economic stability.

What should agencies and enterprises do about the Executive Order?

President Biden's Executive Order underlines the pressing need to ensure the responsible and secure deployment of AI across various sectors. Federal agencies and enterprises embarking in the development of AI systems are encouraged to utilize privacy-enhancing technologies (PETs) to protect the privacy and integrity of data and code with which they operate. In that sense, Confidential Computing should be evaluated as a key enabling technology in achieving the objectives outlined in the Order, by providing trusted execution environments for AI development and deployment. 

As the world continues to embrace the transformative power of AI, the incorporation of Confidential Computing technology stands as an imperative step towards building a safer and equitable digital future. To learn more about how Confidential Computing can bolster your AI endeavors, read our white paper or join our live demo.