When AI comes knocking, how governments can prepare

Today, calling the government often means being greeted by a robotic voice, dealing with a confusing phone menu, and starting over each time you’re transferred to a different department. Online options aren’t much better where complex submission requirements and written documentation have historically suppressed demand. Faced with antiquated, frustrating systems, citizens are increasingly turning to AI agents to handle these interactions on their behalf. 

Most governments aren't yet equipped to handle this shift in volume, and the pressure will only grow as more citizens adopt personal agents. 

To help governments deal with this surge, we released synthetic voice detection in ElevenAgents. It picks up traces in a live audio stream and can tell whether the caller is human or AI. You can set rules to handle these calls, such as moving human callers to the front of the queue, sending AI callers to a dedicated agent with specific instructions and limits, and preventing sensitive actions (like a password reset, or account access) from happening. 

The surge is already here

Government agencies are already seeing sharp increases in requests. Researchers tracked 84 cases across 11 countries where public services faced surges in filings, appeals, and complaints, with officials increasingly pointing to AI as the cause. The researchers call this “agentic flooding”: an AI-driven rise in the volume or complexity of requests that strains an organization’s capacity. Because budget constraints leave many government services running near capacity, even a modest rise in demand can push them past it.

Capacity beats friction

Governments facing this surge have two options. They can add friction through fees, IP blocks, and stricter forms. These can work, but also decreases the accessibility of the services for the people who need them the most. The other option is for them to add capacity, meeting agents with agents.

Capacity is the only option that makes sense in the long term because it can significantly reduce administrative burden and increase throughput. And the organizations that build it will be rewarded with more signal, leading to a higher-resolution picture of what their citizens actually want and need.

Governments can take proactive measures now to mitigate the risks of agentic flooding: 

1. Identify the most vulnerable services based on the complexity of their submission process and the incentive for users to complete it. 
2. Deploy AI agents that can detect and manage AI callers, starting with simple use cases like sentiment analysis of public comments, or fraud detection and prevention. 
3. Expand agent coverage to more use cases, like intake screening, triage, automated correspondence. 

Take the city of Midland, Texas, which serves 138,000 residents and receives 3,000 calls a day. A good portion of those calls went unanswered. With ElevenAgents, the city was able to build Jacky, the AI civic concierge for Midland. Jacky is always available to answer commonly asked questions and support the city’s multilingual population. The city also introduced a chat function on their website, powered by ElevenLabs, to further reduce the volume of phone calls. 

In the Czech Republic, ElevenLabs supports national benefits and employment hotlines with AI agents that handle roughly 5,000 calls per day. These agents independently resolve around 85% of calls, while dramatically increasing answer rates from about 60% to near-universal coverage.

If you’d like to learn more about synthetic voice detection and how ElevenAgents can help you mitigate the risks of agentic flooding and improve public services, we’d love to talk.