A new approach to AI governance

Generative AI is here and to help mitigate risks, simplify implementation, and take advantage of opportunity, federal agencies should establish an enterprise-wide approach and governance framework for AI adoption. Both can help agencies align their AI efforts with their data strategy, respond to evolving federal policies and regulations and support successful agency-wide AI adoption.

As federal agencies assess mission workloads and look to add generative AI to their toolkits, a governance framework will also be crucial to mitigating the risks of AI and harnessing its benefits. In fact, and in line with the implementation of President Biden’s AI executive order, the Office of Management and Budget (OMB) recently released its first policy aimed at doing just that. The new OMB policy announced by Vice President Harris classifies specific actions into three categories: strengthening AI governance, advancing responsible AI innovation and managing risks from the use of AI.

“AI systems are enabling important advancements for federal agencies,” said Vanessa Hunt, general manager, technology, U.S. federal market at IBM. “More traditional task specific AI has been driving operational efficiency, productivity and decision-making for over a decade. With the emergence of generative AI, agencies can take AI to the next level.”

For example, working with IBM, NASA is leveraging the power of foundational models to make weather and climate predictions more accurate and accessible. But even as agencies explore new AI opportunities that exploration must be secure, transparent and responsible. The new OMB policy also addresses potential generative AI experimentation by helping eliminate unnecessary barriers to responsible AI within agencies and encouraging them to responsibly experiment with the technology.

IBM advances ethical AI and governance through watsonx.governance™, a component of watsonx™, a data and AI platform designed to help agencies monitor and govern the entire AI lifecycle. 

By automating risk management and governance, government agencies can accelerate responsible, transparent, and explainable AI workflows, tapping into the power of their data and opening the door to experimentation and identification of new mission customized opportunities – all while also addressing OMB’s new government-wide AI policy.