Wisevoter
Vermont

Vermont Tested Artificial Intelligence Systems

The state government entered a six-month, no-cost agreement with Anthropic to test cybersecurity infrastructure.

Updated on Sept. 30, 2026 in Artificial Intelligence

Bold flat-color editorial illustration featuring a geometric server construct in navy, cream, and red, representing cybersecurity infrastructure.
Vermont has partnered with Anthropic in a six-month, no-cost pilot to test the security of state digital infrastructure against potential cyber threats. AI Illustration. Upload story photo >

Live Poll

Should government agencies provide more transparency when testing new technologies like artificial intelligence?

Vermont officials have disclosed a six-month agreement with AI firm Anthropic to evaluate state cybersecurity systems. The initiative focuses on pressure-testing IT environments to prepare for potential cyberattacks.

Why it matters

Using AI for vulnerability detection allows the state to simulate cyberattacks and strengthen defenses proactively. This approach aims to identify security gaps in government infrastructure without incurring additional costs.

The agency utilizes AI within test and development environments for cybersecurity policy pressure-testing. The agreement with the Anthropic State, Local, Tribal and Territorial program concludes in December 2026.

The players

Vermont Agency of Digital Services

This department manages the information technology infrastructure and digital security policies for the state government.

Anthropic

This artificial intelligence company develops large language models and security-focused software solutions for public and private organizations.

Laura Sibilia

She is a member of the Vermont House of Representatives who has questioned state government transparency regarding tech partnerships.

The details

The Vermont Agency of Digital Services integrated Anthropic tools to identify vulnerabilities within its existing IT systems. Lawmakers examined the project after Rep. Laura Sibilia discovered the partnership during a tech summit.

Timeline

  1. The agreement with Anthropic was executed in June 2026.

  2. Rep. Laura Sibilia learned of the agreement in early August 2026.

  3. Lawmakers held a testimony hearing on September 30, 2026.

  4. The AI evaluation period is scheduled to end in December 2026.

The Tech Race

The state's shift toward AI-driven security mirrors a broader move by public entities to adopt advanced computational tools for threat detection. This strategy replaces manual, periodic scanning with continuous, automated systems designed to keep pace with evolving cyber threats.

The project aims to secure sensitive state IT systems, potentially preventing service interruptions for citizens accessing government databases. Residents will not experience direct changes to their interactions with state digital portals as the technology is confined to internal testing.

The takeaway

Proactive cybersecurity testing using artificial intelligence helps state agencies identify and resolve system vulnerabilities before they are exploited. Effective digital defense requires a balance between rapid technological adoption and rigorous legislative oversight.

What happens next

The evaluation of the AI cybersecurity program will conclude in December 2026, at which point the state may determine future procurement steps.

Further reading

Learn more about local policy by visiting the Artificial Intelligence section.

Source note: This article includes information reported by My NBC5.

Live Poll

Should government agencies provide more transparency when testing new technologies like artificial intelligence?