DOD TEVV standards will establish a global model – this promotes interoperability and reduces our vulnerabilities
Flournoy, Haines et al 2020 - former Under Secretary of Defense for Policy and Director of National Intelligence [Michèle and Avril, Gabrielle Cheftiz Special Assistant to the Under Secretary of Defense for Policy October, “Building Trust through Testing Adapting DOD’s Test & Evaluation, Validation & Verification (TEVV) Enterprise for Machine Learning Systems, including Deep Learning Systems” https://cset. georgetown.edu/wp-content/uploads/Building-Trust-Through-Testing.pdf Acc 6/23/22 JZ]
7. Develop industry/U.S. government TEVV standards and promote them internationally. DoD, working with the National Institute of Standards and Technology (NIST) and industry, and building on DoD requirements and processes developed for industry, should develop standards for ML/DL testing for the private sector that can be publicly promoted and help inform private sector development of ML/DL systems. Such standards would focus on a range of issues, including robustness, interpretability, performance metrics, fail-safe design, traceability for data collection and management, privacy, and testability. These should be broad standards that serve as guidance for both government and commercial developers to develop operationally specific design requirements and testing metrics. DoD entities—including the JAIC and R&E—and the IC are already playing a role in the development of U.S. government AI standards, led by NIST and the Office of Science and Technology Policy. NIST’s 2019 plan on U.S. leadership on AI provides an important foundation and calls for the development of standards and metrics for trustworthiness (e.g., accuracy, explainability, resiliency, safety, reliability, objectivity, and security), complexity, domain-specific and context-dependent risk, and uncertainty.35 NIST and the State Department should play a leading role in promoting U.S. government standards for AI and ML development and testing domestically and throughout international standards-setting organizations, such as the International Standards Organization and Institute for Electrical and Electronics Engineering, and multilateral institutions, such as the OECD AI Policy Observatory. The promotion of U.S. standards globally will help bolster U.S. economic competitiveness, create a level playing field for U.S. industries who are collaborating with DoD and consequently subject to these standards, incorporate U.S. values and ethical principles into AI and ML development, and ensure that the United States and its allies are interoperable. As we’ve seen in other critical technology areas, the United States must ensure that competitors do not set standards, which make it harder to manage vulnerabilities and hinder U.S. efforts to establish the highest degree of ethics, safety, and risk management.