The U.S. Environmental Protection Agency (EPA) is facing growing cybersecurity and data management risks that could undermine the reliability of environmental data and hinder the agency’s artificial intelligence ambitions, according to a new Office of Inspector General (OIG) report.
Central to the issue is the EPA’s Central Data Exchange, an online portal used by regulated entities to submit environmental information. The OIG flagged vulnerabilities in the system, citing unverified users, inactive accounts, and questionable data quality.
The EPA manages more than 41,500 terabytes of environmental data, raising concerns over storage, access, and governance, particularly after workforce reductions eliminated nearly 23% of the agency’s staff last year. As scientific personnel were reassigned, questions emerged about the ownership and stewardship of critical datasets.
The OIG also warned that weak data oversight could hamper the EPA’s AI deployment efforts, urging stronger governance, inventory management, and testing protocols. Recommendations included continuous monitoring of AI systems, staff training, impact assessments, and public feedback mechanisms.
“It is important for the EPA to develop and implement strategies that effectively address the human health and environmental challenges posed by the growing frequency and costs of natural and manmade disasters. The EPA may also need to enhance its disaster-response capabilities to better protect people and the environment from these increasing threats and to ensure that it is prepared to handle future disasters effectively,” the watchdog said.
In response, the EPA said it is modernizing legacy systems, strengthening cybersecurity measures, and establishing governance frameworks to support secure AI adoption while safeguarding scientific data integrity.