A new report from the U.S. Government Accountability Office (GAO) is sounding the alarm on the environmental and societal risks posed by generative AI (GenAI) technologies — calling for greater transparency and stronger oversight from policymakers.

GAO outlines a wide range of potential threats tied to GenAI, including high energy and water consumption, workforce disruption, misinformation, cybersecurity vulnerabilities, data privacy concerns and broader national security implications.

“Generative AI uses significant energy and water resources, but companies are generally not reporting details of these uses,” the report states. Most current estimates focus on the carbon emissions and electricity required to train AI models but data on water use and full lifecycle environmental impact remains sparse.

The report cites projections that U.S. data centers that are already responsible for roughly 4% of national electricity consumption in 2022 could see that number rise to 6% by 2026, with AI workloads being a major contributor.

Beyond environmental impacts, the GAO highlights five key human-centered risks:

• Lack of accountability
• Unintentional bias
• Unsafe or unverified systems
• Cybersecurity threats
• Inadequate data privacy protections

To mitigate these risks, the GAO recommends that federal agencies and industry leaders collaborate to improve data reporting practices, invest in eco-innovative solutions, adopt responsible AI frameworks, and share best practices across sectors.

While policymakers could opt to maintain the status quo, the GAO warns that such an approach may not be sufficient to address the complex and evolving risks presented by GenAI.