The Office of Management and Budget (OMB) released a draft memorandum that provides guidance to federal agencies on how to develop regulatory and non-regulatory approaches to the use and development of artificial intelligence (AI).
On January 7, 2020, the Office of Management and Budget (OMB) released a draft memorandum that provides guidance to federal agencies on how to:
Develop regulatory and non-regulatory approaches regarding artificial intelligence (AI) technologies.
Consider ways to reduce barriers to the use of AI in order to promote innovation.
The guidance sets out policy considerations for the use and development of AI specifically in the private sector. The guidance also provides the following ten principles to consider when formulating regulatory and non-regulatory approaches to the development and use of AI:
Public trust in AI. Establish public trust by promoting reliable AI applications that address risks to privacy and individual rights.
Public participation. Encourage public participation in the rulemaking process and promote public awareness of AI standards and technology.
Scientific Integrity and Information Quality. Leverage scientific and technical information and processes by holding information that is expected to have a substantial influence on public policy or private sector decisions to high standards of quality, transparency, and compliance.
Risk Assessment and Management. Use a transparent risk assessment and risk management approach consistently across agencies and technologies to determine whether risks are acceptable.
Benefits and Costs. Assess the full societal costs, benefits, and distributional effects associated with the development and deployment of AI.
Flexibility. Pursue performance-based and flexible approaches that can adapt to rapid changes and updates to AI.
Fairness and Non-Discrimination. Evaluate issues of fairness and non-discrimination and whether the AI application:
reduces levels of unlawful, unfair, or otherwise unintended discrimination compared to existing processes; or
introduces real-world bias that produces discriminatory outcomes.
Disclosure and Transparency. Determine appropriate levels of transparency and disclosure of the use of AI to increase public trust.
Safety and Security. Keep controls in place to ensure the confidentiality, integrity, and availability of the information processed, stored, and transmitted by AI and promote the development of AI that is safe and secure.
Interagency Coordination. Coordinate with other agencies to ensure consistency and predictability of AI-related policies.
The guidance provides that if existing regulations are sufficient or the benefits of a new regulation do not justify the cost, agencies may consider either not taking any action or identifying non-regulatory approaches that may be appropriate, such as:
Sector-specific policy guidelines or frameworks to encourage AI innovation.
Granting exemptions or allowing pilot programs that provide safe harbors for specific AI applications.
Voluntary consensus standards for AI applications.
Implementing agencies with regulatory authorities are required to review and submit plans to the OMB on how the agency will achieve consistency with the memorandum.