AI agents are flooding public services with new requests

In a significant development reshaping how citizens interact with government bodies, Artificial Intelligence (AI) agents are reportedly causing a substantial increase in requests made to public services worldwide. This phenomenon, which researchers are calling "agentic flooding," highlights both the growing accessibility of AI tools and the evolving landscape of public administration.

A recent analysis, as highlighted by TechCrunch, reveals that the vast majority of these AI-assisted submissions come from individuals legitimately seeking benefits or services they are entitled to receive. This suggests that AI is empowering citizens to overcome bureaucratic hurdles, but it simultaneously presents a new challenge for public sector capacity.

The Rise of Agentic Flooding in Public Services

The term "agentic flooding" describes a surge in the volume or complexity of requests to a service caused by the use of AI. This trend has been identified by researchers Chris Schmitz of the Hertie School, and Lewis Hammond and Alan Chan of the Centre for the Governance of AI. Their analysis, which examined 84 potential cases across 11 jurisdictions, indicates a sharp growth in public-service requests that aligns with the wider availability of AI tools since 2022.

What exactly does this mean? It means that ordinary people are increasingly turning to AI to help them with tasks that were previously daunting or time-consuming. Imagine a complex application form for a public benefit, a detailed complaint against a service provider, or even a judicial appeal. These processes often involve intricate language, specific legal requirements, and multiple steps. For many, navigating such systems can be a significant barrier. AI agents, however, can simplify this. By allowing users to paste or even photograph official documents, these tools can interpret the content, gather necessary context, and help draft a coherent, usable response or application.

The impact is tangible. For instance, the Consumer Financial Protection Bureau reportedly saw complaints increase fivefold over a period where AI tools became more accessible. Similar patterns have been observed in Brazilian judicial petitions, German parliamentary petitions, Australian Freedom of Information (FOI) requests, and specific UK public service channels like Money Claims Online and local plan consultations. This broad geographical and functional spread suggests a systemic shift rather than isolated incidents.

How AI Agents Are Empowering Citizens

The underlying mechanism is straightforward yet powerful. AI agents, by definition, are AI systems designed to act autonomously, making decisions and taking actions to achieve defined goals without constant human oversight. In the context of public services, they function as sophisticated digital assistants. They can help individuals understand official letters, collect required information, and assemble comprehensive submissions that might have otherwise been abandoned due to complexity or difficulty.

This capability is particularly beneficial for services that accept free text of unlimited length through open digital channels, such as FOI portals or public consultation platforms. These are the areas most exposed to agentic flooding, as AI can generate detailed, lengthy, and often well-structured submissions that mimic human-authored content.

The core finding, that "the vast majority of cases we find are people who are entitled to claim for something, claiming for that thing," is crucial. It indicates that AI is not primarily being used for fraudulent or malicious purposes in this context, but rather as a tool for improved access and equity. It allows individuals who might lack the time, expertise, or confidence to navigate bureaucracy to effectively assert their rights and access services.

Implications for Public Services and Governments

While empowering citizens is a positive outcome, the sudden increase in request volume presents a significant operational challenge for public authorities. Many government agencies are not equipped to handle such a rapid escalation in inbound communications. The researchers warn that public authorities might soon run out of capacity if they do not adapt quickly.

One potential response from governments, as suggested by the research, could be the reintroduction of application fees or the implementation of other barriers to manage the influx, rather than investing in increased capacity. Australia's federal FOI regime, for example, has considered reintroducing fees after officials noted a wave of AI-generated requests. This approach, however, could counteract the positive impact of AI in improving access for entitled individuals.

The challenge for governments is multifaceted:

• Capacity Strain: Increased volume means more work for human staff who must still assess each case, regardless of whether it was AI-assisted.
• System Redesign: Agencies may need to rethink their application systems to better integrate with or manage AI-assisted claimants, ensuring fair access without overwhelming staff.
• Ethical Considerations: While AI aids legitimate claims, governments must also be vigilant against potential misuse, such as the generation of false claims, which some research indicates is a risk if AI handles processes like billing or compliance without human review.
• Investment in AI: To counter the "flooding," governments themselves may need to accelerate their adoption of AI to automate internal processes, manage inbound requests, and improve overall efficiency.

The Broader Trend: AI in Government Operations

The phenomenon of "agentic flooding" is occurring within a broader global trend of governments integrating AI into their operations. The U.S. federal government alone reported over 1,700 AI use cases by the end of 2024, more than doubling its previous count. Countries like Estonia are experimenting with AI in numerous public administrations, and regions like the Middle East and Africa are seeing booming conversational AI markets for customer service.

Major tech companies are actively partnering with governments to facilitate this transition. For example, OpenAI recently secured a new "OneGov" deal with the General Services Administration (GSA) to provide ChatGPT models to federal government employees through 2028. This deal aims to enhance federal performance, reduce administrative burden, and allow agencies to focus more on delivering results to the American people. Federal agencies will receive discounted, consumption-based pricing for ChatGPT models, including those in FedRAMP-authorized environments.

AI agents are poised to transform public service delivery, moving beyond simple chatbots to proactive, goal-driven systems. These advanced agents can automate complex, multi-step workflows, from verifying permit applications to coordinating disaster response logistics. They promise 24/7 citizen services and the ability to offer personalized, proactive support, such as notifying families about benefits they qualify for and assisting with applications.

However, this rapid integration also brings challenges that need careful consideration. The Alan Turing Institute, for instance, focuses on the responsible and effective adoption of AI by government, conducting research into its application, integration, and oversight. Ensuring transparent data practices, accountable human-AI collaboration, and managing public perceptions of AI are critical for successful implementation.

The Path Forward

The "agentic flooding" of public services underscores a pivotal moment in the intersection of AI and governance. While it highlights the power of AI to democratize access to public services and empower citizens, it also exposes the need for governments to rapidly adapt their infrastructure and policies.

Instead of simply creating barriers, a more sustainable approach might involve governments leveraging AI themselves to process the increased volume, streamline internal workflows, and develop more resilient and responsive public service systems. This could mean deploying their own AI agents to triage requests, automate routine responses, or even proactively identify eligible citizens for services. The ultimate goal should be to harness AI's potential to create more efficient, transparent, and citizen-centric public administration, ensuring that the benefits of AI are realized by both the public and the institutions serving them.

Frequently Asked Questions

What is "agentic flooding" in the context of public services?

"Agentic flooding" refers to a significant increase in the volume or complexity of requests, claims, or applications submitted to public services, directly caused by individuals using AI agents or tools to assist them in preparing these submissions.

Who are the researchers behind the concept of "agentic flooding"?

The concept of "agentic flooding" and the associated research have been put forward by Chris Schmitz of the Hertie School, along with Lewis Hammond and Alan Chan of the Centre for the Governance of AI.

Are these AI-generated requests typically legitimate?

Yes, according to the researchers, "the vast majority of cases we find are people who are entitled to claim for something, claiming for that thing." This suggests AI is largely helping individuals access services they are already eligible for.

How are governments responding to the increased volume of AI-assisted requests?

Governments are currently facing challenges in handling the increased volume. Potential responses include considering the reintroduction of application fees or other barriers, but there is also a growing trend of governments themselves adopting AI to improve efficiency and manage services.