Efficient Government and Safe Innovation: A Collaborative Approach to Artificial Intelligence Policy

By: Greg Ferenstein (Brookings Institution/Tech Tank)

In this commentary piece, author Greg Ferenstein shares his view on how artificial intelligence can enhance government efficiency and regulatory cooperation.

In 2019, a small typo in a California man’s tax document led to his refund being sent to a stranger. While the Internal Revenue Service (IRS) acknowledged the mistake, California’s financial agency did not, leaving him unable to recover his money, as reported by an ABC affiliate.

Now, imagine if errors like these—common in government paperwork—could be prevented by intelligent software designed to handle routine tasks with precision. Advanced generative AI (Gen AI), such as ChatGPT, has the potential to streamline bureaucratic processes, catch costly mistakes, and alleviate administrative burdens for both citizens and agencies. Recognizing this, OpenAI recently introduced ChatGPT Gov, a version tailored for public sector use.

Federal agencies should be encouraged to explore AI-driven solutions within their operations.

Businesses leveraging AI have experienced significant productivity gains, and the same technology could simplify interactions with government services, cutting down on inefficiencies and reducing the staffing required to handle avoidable bureaucratic hurdles.

President Trump’s Department on Government Efficiency (DOGE) has aggressively pursued cost-cutting measures, including workforce reductions, a move that has drawn criticism from congressional Democrats. While government efficiency became a politically charged issue during the 2024 election, AI presents a rare opportunity for bipartisan progress. Even President Obama championed waste reduction initiatives early in his administration.

Beyond improving public sector efficiency, expanding AI adoption in government could also foster a more cooperative relationship between regulators and the tech industry.

By integrating AI into their own workflows, agencies could help shape global compliance standards for software developers navigating the increasingly complex landscape of safety regulations across different jurisdictions.

While only a handful of companies—such as OpenAI and Google—have the resources to create foundational AI models, fine-tuning these models for specific governmental needs can be done at minimal cost. Regulations like the EU AI Act and various U.S. state proposals introduce legal complexities that require AI systems, including modified models, to comply with strict ethical and security guidelines.

Crucially, these regulations don’t just apply to private companies—they also govern government AI use…

CONTINUE READING…