Maine Dispatchers Cautiously Adopt AI for Non-Emergency Calls

Maine’s 911 dispatch centers are grappling with a persistent crisis: call volumes keep rising while staff levels remain critically low. To address this imbalance, several agencies are introducing artificial intelligence to handle non-emergency inquiries. However, the technology is being used with strict limitations, ensuring that AI never touches the most critical, life-threatening calls.

The pressure is most evident in Penobscot County, where the Regional Communications Center has led the state in call volume for six consecutive years. Despite having fewer than half of its positions filled, the center manages over 60,000 emergency calls annually, alongside hundreds of thousands of non-urgent requests. This workload creates significant strain on dispatchers, who must quickly triage every incoming call as a potential emergency.

AI Handles Routine Inquiries

To alleviate this burden, the Penobscot County center implemented an AI-powered system for its non-emergency line. This tool acts as a digital phone tree, offering callers options for issues that do not require immediate police, fire, or medical response. It can direct users to online reporting tools or connect them to a human dispatcher if the situation escalates, but it does not answer 911 emergency calls.

Director Brad Fitzgerald explains that the system is designed to keep dispatchers available for genuine emergencies. While the AI does not eliminate the volume of non-urgent calls entirely, it absorbs a small but meaningful percentage of routine inquiries. This allows staff to focus their limited energy on high-stakes situations rather than answering questions about fireworks schedules or minor municipal issues.

Statewide Adoption Remains Slow

Despite the potential benefits, AI adoption across Maine’s dispatch centers is proceeding cautiously. A survey by GN technics/ai (en-US) found that only two agencies are currently using the technology, while three others plan to launch similar systems within the next year. Most directors acknowledge the value of automating routine tasks but are hesitant to commit until the technology has a stronger track record of reliability.

Cost and infrastructure upgrades also play a role in the slow rollout. Agencies are weighing the financial investment against the immediate operational gains. The prevailing sentiment is one of careful experimentation rather than full-scale deployment, reflecting a broader hesitation in public safety sectors to rely on new technology for critical infrastructure.

Human Judgment Stays Central

The primary trade-off in this approach is the manual oversight required to maintain trust and accuracy. Agencies are not using AI to make final decisions on emergency responses; instead, it serves as a filter to reduce noise in the system. This ensures that human dispatchers retain full control over the most vital interactions, preserving the accountability and empathy necessary for public safety operations.