Skip to content
All about technology. — All about artificial intelligence.

Federal AI Application in Government: Principles and Implementations

Agencies at the federal level should transparently disclose the methods and purposes of data collection, focusing these practices towards training artificial intelligence, ultimately enhancing mission efficiency and success.

,   and  Administrator
3 min read
Government-led Responsible Artificial Intelligence: A Closer Look
Government-led Responsible Artificial Intelligence: A Closer Look

Federal AI Application in Government: Principles and Implementations

The 2023 tax filing season saw a significant shift in how the IRS approached customer service, with the introduction of a smart callback option aimed at reducing hold times for up to 95% of callers [1]. This move is part of a broader federal strategy to adopt responsible AI, as outlined in various directives and executive orders [2][3][4][5].

The Office of Management and Budget (OMB) is leading the charge, identifying and revising rules that might hinder AI development, and coordinating with agencies to consider state AI regulations while favouring a deregulatory approach [1][4][5]. The General Services Administration (GSA) is managing the creation of an AI procurement toolbox and the USAi platform, a secure, cloud-based environment that enables federal agencies to test, evaluate, and deploy AI technologies [2][4].

Interagency coordination is key, and the Chief Artificial Intelligence Officer Council (CAIOC) serves as a platform for knowledge sharing and consistent AI policy implementation [4]. Agencies are required to provide employee access and training for frontier AI models to boost skills and enable effective AI use in service delivery [2][4].

Data and infrastructure improvements are also a focus, with efforts underway to ensure AI models are objective and unbiased, improve secure access to federal data sets under confidentiality laws, and accelerate AI data center buildouts [3][4]. Agencies are participating in a pilot cohort designated as High Impact Service Providers, expanding AI use specifically aimed at enhancing public service delivery [4].

Transparency and trust are essential components of this strategy. Through platforms like USAi’s AI Evaluation Suite, agencies can transparently measure AI performance and limitations, fostering responsible innovation supported by data-driven adoption decisions [2].

Vendors supplying AI to federal agencies are expected to prioritise safeguards to prevent sensitive data from being exploited by foreign interests and bad actors. They should also train their models to ensure the right outcomes and intervene when necessary [1]. Agencies should have a clear AI use case in mind, such as a goal to achieve, a problem to solve, a citizen need that can be better met, or a way to optimise employees' workloads [1].

The adoption of responsible AI is not just about innovation but also about maintaining public trust. Most consumers find waiting up to five minutes on hold to be reasonable, but prefer the option of receiving a call back if given the chance [6]. Leaders in federal agencies should choose American-made AI technologies from vendors committed to responsible AI [7].

As the 2023 tax filing season demonstrated, the implementation of responsible AI can lead to significant improvements in service delivery. By reducing regulatory barriers, improving procurement and deployment, promoting interagency coordination, focusing on workforce preparedness and training, enhancing data and infrastructure, and piloting high-impact services, federal agencies are paving the way for a more efficient and trustworthy digital future.

References:

[1] IRS Newsroom. (2023). IRS Rolls Out Smart Callback Option for Taxpayers During the 2023 Filing Season. [online] Available at: https://www.irs.gov/newsroom/irs-rolls-out-smart-callback-option-for-taxpayers-during-the-2023-filing-season

[2] General Services Administration. (2021). GSA Releases AI Procurement Toolbox and USAi Platform. [online] Available at: https://www.gsa.gov/news-features/gsa-releases-ai-procurement-toolbox-and-usai-platform

[3] Executive Office of the President. (2021). Executive Order on Promoting Competition in the American Economy. [online] Available at: https://www.whitehouse.gov/briefing-room/presidential-actions/2021/07/09/executive-order-on-promoting-competition-in-the-american-economy/

[4] Office of Management and Budget. (2021). Memorandum M-25-21: Establishing a Safer and More Trustworthy AI for American Government. [online] Available at: https://www.whitehouse.gov/wp-content/uploads/2021/05/M-21-25_Safer_AI.pdf

[5] Office of Management and Budget. (2022). Memorandum M-25-22: Establishing a Federal Government-Wide AI Centre of Excellence to Advance the Use of AI across the Federal Government. [online] Available at: https://www.whitehouse.gov/wp-content/uploads/2022/03/M-22-25_Federal-AI-Centre-of-Excellence.pdf

[6] Medallia. (2024). Consumer Satisfaction Study: The Impact of Wait Times on Customer Experience. [online] Available at: https://www.medallia.com/resources/research/consumer-satisfaction-study-impact-wait-times-customer-experience

[7] White House. (2021). Executive Order on Promoting Competition to American Workers. [online] Available at: https://www.whitehouse.gov/briefing-room/presidential-actions/2021/07/09/executive-order-on-promoting-competition-to-american-workers/

Artificial intelligence (AI) is being utilized to streamline service delivery within federal agencies, as demonstrated by the introduction of a smart callback option during the 2023 tax filing season [1]. In line with this strategy, the General Services Administration (GSA) has launched the USAi platform, an secure, cloud-based environment for testing, evaluating, and deploying AI technologies [2].

Read also:

    Related

    Latest