GAO Releases Accountability Framework on AI
In June 2021, the Government Accountability Office (GAO) released an accountability framework on artificial intelligence (AI) for agencies to follow as they incorporate new and emerging technologies into their day-to-day operations.
The framework explains, “GAO’s objective was to identify key practices to help ensure accountability and responsible AI use by federal agencies and other entities involved in the design, development, deployment, and continuous monitoring of AI systems.”
A federal framework has long been anticipated as agency inspectors general and legal counsel find ways to make credible assessments and decisions surrounding AI. Lawmakers have also expressed concern regarding the lack of a unified national strategy for handling AI.
Taka Ariga, Chief Data Scientist at GAO, explained to FedScoop, “The only way for you to verify that your AI, in fact, is not biased is through independent verification, and that piece of the conversation has been largely missing. So GAO, given our oversight role, decided to take a proactive step in filling that gap and not necessarily wait for some technology maturity plateau before we addressed it.”
By taking a proactive measure to address the use of AI in the federal government, GAO is working to ensure that no agencies fall behind as AI becomes more advanced. The framework is organized around four principles: governance, data, performance, and monitoring.
The framework addresses many complex issues that surround AI like establishing a standard risk-management framework that protects privacy. Another goal of the framework is to establish a human aspect to AI use in the federal government by including human supervision and oversight. GAO’s new framework adds requirements, documentation and evaluation to the realm of AI and incorporates it in government functions.
The framework includes information on the phases of AI development. The coming third-wave of AI systems develop contextual adaptation, where AI models explain their decisions to drive further decisions.
The Department of Defense has attempted to audit AI in the past, but, according to FedScoop, “DOD ran into issues because there was no standard definition of AI, and it lacked AI inventories to assess.” GAO has already begun using its new framework to launch investigations into AI use. In the future, the government will work to create ethical standards for the use of AI.
Ariga continued, “Maybe not tomorrow but certainly in the near future because, at least in the public sector domain, our responsibility to get it right is so high. A lot of these AI implementations actually do have life or death consequences.”
