As artificial intelligence becomes more common, so do concerns about what could happen if the technology is not properly controlled.
Recent reports of AI agents accessing systems on the open internet in September have raised new questions among lawmakers about how the technology is tested and kept in check. Those concerns are now reaching local government, where officials are weighing the benefits of AI against the risks of giving it access to sensitive information.
In Olmsted County, employees can use approved AI tools, but they are prohibited from entering private or confidential information. The county also has technical safeguards and human oversight in place to help keep sensitive data from being sent to AI systems.
“We only use artificial intelligence models that we can control or know where our data is being policed or used,” said Michael Stier, Olmsted County’s information technology security architect.
Stier said the county generally uses AI systems that operate within the continental U.S. so the county knows where its data is being stored and to make sure vendors are following its privacy requirements.
Olmsted has been ready…
The county has had an AI policy for about three years, and about two years ago created an AI governance group to help oversee how the technology is used.
The group has about a dozen members and works with the county’s information technology and security teams. The county is now working to expand that effort to include more areas of county government.
The article outlines national and local benefits and concerns…
Despite those concerns, the county is continuing to explore ways AI could help employees and community members.
Stier said the county is testing potential uses that could help residents find the right county services, as well as tools that could help people navigate the county website.
Any new use would still go through the county’s security and governance process