AI ethics in public administration
In the ongoing quest to improve government services and meet evolving mission demands, federal agencies are increasingly turning to Artificial Intelligence (AI) to streamline processes, reduce bureaucratic bottlenecks, and enhance the end-user experience.
One of the first steps towards this transformation is the implementation of AI by the Internal Revenue Service (IRS), with the introduction of a smart callback option to reduce long wait times for callers. Similarly, the Social Security Administration has used AI to address a root cause of calls by adding a "return to my saved application" button.
However, the adoption of AI comes with its own set of challenges. Agencies must ensure that AI models are trained to deliver the right outcomes and have the ability to intervene when necessary. This is particularly important in sensitive areas such as national security and service delivery, where the misuse of AI could have serious consequences.
To address these challenges, federal agencies are required to establish centralized governance for AI, prioritizing transparency and ethics, investing in infrastructure and workforce readiness, and aligning AI strategies with data security and ethical principles.
The establishment of the Chief AI Officer Council (CAIOC) will coordinate AI efforts across agencies, creating and implementing unified governance frameworks that emphasize safety, fairness, transparency, and vendor accountability. This centralized oversight facilitates consistency in ethical standards and risk management across federal AI deployments.
Agencies should also identify priority AI use cases aligned with agency missions, tailoring AI deployment to areas where it offers the most value while managing risk through early regulatory attention and infrastructure investment.
Effective AI implementation requires improving compute infrastructure, ensuring access to high-quality, well-governed datasets to prevent poor performance and compliance issues, and considering the use of open-source AI models with caution to balance innovation with traceability and misuse risk.
Building an AI-ready federal workforce that is knowledgeable about AI’s capabilities, limitations, and ethical use is crucial. This includes continuous training, creating communities of practice for sharing best practices, and emphasizing human-centric and responsible oversight of AI systems.
AI systems must be “purposeful and performance-driven,” with traceability and regular monitoring. Agencies must ensure AI use respects privacy, fairness, and security, aligning with executive orders mandating responsible AI development and use as a national priority.
Risk management and security focus are also essential. Agencies need to address emerging risks including cybersecurity, biosecurity, and national security concerns. This involves implementing risk controls, incident reporting systems, and compliance with procurement policies that emphasize the use of objective, secure, and reliable AI systems.
Furthermore, agencies should be clear about how data is being collected and used to train AI models. Adopting AI in government aims to enable transformation, modernization, and create more efficiency. A 2024 Medallia study found that shorter wait times and call back options are highly valued by consumers, with 58% finding waiting up to five minutes on hold reasonable, but 66% preferring call back options.
AI can also be used to detect citizen feedback related to scams and notify the appropriate teams. For instance, a state DMV is looking into using AI to detect when users are experiencing issues with their website and proactively trigger outbound contact center outreach via smart scheduling tools.
In conclusion, by adopting a thoughtful, mission-driven approach to responsible AI, federal agencies can effectively implement AI technologies to improve government services while protecting citizen data and upholding ethical standards in line with evolving federal AI policies and action plans. Agencies are also required to leverage commercial off-the-shelf American-made AI solutions designed to accelerate federal efficiency and improve self-service. Choosing technology partners with safeguards to prevent sensitive data exploitation is essential in this regard.
- To ensure the federal workforce is ready to harness the capabilities of Artificial Intelligence (AI), it's vital to provide continuous training and establish communities of practice, fostering a workforce reimagined with the knowledge and ethics necessary for responsible AI usage.
- As federal agencies pursue the implementation of AI to streamline processes and enhance services, they must be mindful of the potential risks associated with the technology, prioritizing the establishment of centralized governance that focuses on safety, fairness, and transparency, thereby aligning with federal AI policies and action plans.
