US Executive Order on Artificial Intelligence Demands Safe, Secure, and Trustworthy Development
The Biden Administration recently issued an executive order to ensure the safe, ethical, and responsible development of artificial intelligence (AI) technologies. The Order prioritizes a coordinated federal government approach, outlining principles and guidelines to mitigate risks from irresponsible AI systems, while simultaneously maintaining support for innovation that can solve significant challenges.
The Executive Order on Artificial Intelligence is guided by eight core principles:
- The safety and security of AI
- Supporting responsible competition
- Assisting workers in AI-driven transitions
- Ensuring that AI does not conflict with civil rights protections
- Protecting consumers and patients
- Prioritizing privacy and liberties
- Managing federal AI risks
- Advancing American leadership in setting global AI standards
The Order also directs the National Institute of Standards and Technology (NIST) to produce additional guidelines to support the current AI Risk Management Framework (NIST AI 100-1) and take actions to ensure the secure and trustworthy development and use of AI, which will impact companies working with AI technologies.
Companies should pay attention to this executive order to ensure compliance with the new standards and guidelines for AI development and use. Following these guidelines proactively, even if not yet legally required, enables companies to mitigate risks, address biases, and build trust with consumers.
Given AI’s expanding integration and impacts, aligning development and deployment practices with the order’s principles around safety, transparency, and accountability fosters innovation responsibly. Adopting emerging standards will prepare companies for regulatory changes while demonstrating commitments to AI ethics.
Implementation of the Order
The order has introduced a whole-of-government approach, directing specific actions across federal agencies and departments, and establishing new coordinating bodies.
One of the more significant steps is the publication of an AI Bill of Rights, which provides AI safety guidelines. The order also aims at promoting new public-private research partnerships and apprenticeships to assist in developing AI talent. Moreover, the Order calls for anti-discrimination requirements for AI systems used by federal and state programs. Lastly, the advancement of privacy-enhancing technologies and the labeling of AI-generated content are highlighted as key enablers for trustworthy AI adoption.
Equity Assessments and Risk Management
There is a specific direction in the executive order for federal agencies to incorporate equity assessments, risk management practices, and continuous monitoring in all government AI uses that impact Americans’ rights. This includes taking measures such as implementing algorithmic discrimination testing, undertaking enhanced cybersecurity audits, carrying out external red-team vulnerability assessments, and introducing watermarking requirements for AI outputs. Civil rights authorities are also empowered to address any areas of unfair bias or disparity introduced by automated systems.
AI Safety and Security Board
The Order sets up a dedicated AI Safety and Security Board, advising on managing risks from AI integration into vital sectors including finance, healthcare, transportation, and education where there could be significant public impacts.
The order also stresses the importance of supporting American technology leadership and collaboration with international allies in setting global AI standards. It promotes consensus principles and providing technical aid for countries adopting AI in critical infrastructure and burgeoning sectors.
White House AI Council
For execution purposes, a White House AI Council has been established, chaired by the Deputy Chief of Staff. This council includes Cabinet-level heads from all the major federal departments as well as independent agencies.
The council is tasked with coordinating AI policies and guiding the integration of safe AI adoption across various government functions. Overall, the Order presents a comprehensive plan for strengthening both America’s innovative edge in artificial intelligence and the public guardrails required to responsibly reap the benefits of this transformation.
Innovative Cybersecurity Leaders Show Their Support
Kiteworks and other AI Security Innovative Leaders including OpenPolicy, Armis, Cranium.AI, Axonius, HiddenLayer, Protect.AI, Astrix Security, and Cybeats, recently released a joint statement on the Executive Order:
The Executive Order is a welcomed and necessary step forward in ensuring both the trusted deployment of AI and seizing U.S. competitiveness and leadership in AI development. Supporting the trusted deployment and development of AI entails supporting the development of cutting-edge innovative solutions needed to protect government, industry, and society from emerging AI threats. The undersigned are innovative companies of all sizes that stand forefront of developing leading security, safety, and trustworthy AI and privacy solutions. We believe that the Biden-Harris Administration’s Executive Order provides a sound framework for addressing the emerging security and safety AI threats and ensuring that AI is developed and used in a safe, secure, and trustworthy manner.
As an industry leader in content governance, Kiteworks is uniquely positioned to support the Executive Order on AI.
Kiteworks provides robust controls over sensitive content, preventing unauthorized access or use that could lead to data breaches or biased AI training sets. Specifically, Kiteworks allows granular policies to be set protecting content integrity, enforcing access approvals, and controlling downstream usage. These capabilities directly address the order’s aims around privacy preservation and responsible AI adoption. Kiteworks also advocates for security frameworks like the NIST CSF that complement the administration’s push for standardized assessments.
Overall, Kiteworks’ expertise in securing sensitive content against misuse aligns with and enables the Order’s vision of trustworthy AI progress through prudent content governance.
Building Trust in AI Through Prudent Progress
President Biden’s Executive Order on AI represents a crucial step in guiding the responsible development of these transformative technologies. As AI systems become more prevalent, they introduce emerging risks that require coordinated mitigation efforts around biases, privacy breaches, and attacks.
This Order puts into place a strategy which is informed by sector experts like Kiteworks and other leaders to strengthen AI safety through ongoing assessments, audits, and standards implementation. It also aims to accelerate the promise of AI personalizing healthcare, creating safer transport and more engaging education, as well as a higher prioritization of transparency, accountability, and equitable access while balancing innovation. This order works to manage the risks to build widespread trust and enable faster progress of AI done safely.
As AI integration accelerates across industries, Kiteworks provides a robust solution for organizations to ensure regulatory adherence and mitigate emerging content risks. For example, Kiteworks enables robust governance over sensitive data used to train AI systems, directly addressing core tenants of the NIST AI Risk Management Framework.
By allowing granular policies to control access, usage, and visibility of sensitive content, Kiteworks helps organizations mitigate risks around data breaches, biases, and unauthorized exposure that can arise from AI.
Kiteworks empowers demonstrable compliance with the framework’s prescribed practices for continuous monitoring, evaluating algorithmic fairness, and maintaining data integrity. For any entity leveraging AI, Kiteworks provides a scalable path towards responsible oversight of training data while unlocking innovation opportunities.
Whether meeting existing regulations or new safety guidelines, Kiteworks’ expertise in securing sensitive content against misuse provides a reliable path towards the executive order’s vision of trustworthy and equitable AI progress. The Platform’s protections allow entities to fully actualize AI’s potential through a robust data governance approach aligned with emerging best practices.
To learn more how Kiteworks can protect your organization from the risks associated with artificial intelligence, schedule a custom demo today.
Additional Resources
- Blog Post AI for the Good and Bad in Cybersecurity
- Blog Post Building Trust in Generative AI with a Zero Trust Approach
- Blog Post The Promise of DRM and Why It Typically Falls Short
- Webinar The Evolution and Impact of Artificial Intelligence on Data Security Decisions
- Video Kiteworks SafeEDIT: Next-generation Digital Rights Management (DRM)