Agent flood sweeps government portals as study records doubling of complaints and judicial petitions due to smart assistants

A upcoming research paper by researcher Chris Schmitz, to be presented next month at the "AI Ethics and Society" conference, reveals a structural shift in how individuals interact with public administrations, as unprecedented numbers of AI-assisted transactions and formal petitions flow toward government institutions. The study identified 84 cases, which Schmitz labels "agent flood," across 11 different jurisdictions, and showed that generative tools have begun to radically alter the pace of interaction with public services. Figures indicate that complaints to the UK Housing Ombudsman more than doubled after the launch of the ChatGPT model, rising from 2,600 complaints in 2022 to over 7,000 last year, while the US Consumer Financial Protection Bureau recorded a five-fold jump in complaint volume during the same period, with parallel spikes in judicial petitions in Brazil and parliamentary petitions in Germany.

The case data show a uniform pattern in which request volume was almost stable before 2022, then accelerated at increasing rates as the technology spread, without any slowdown in sight.According to Schmitz, this rise is due to declining usage costs and evolving available skills; instead of assembling context and crafting precise, labor-intensive commands as was required with early releases such as ChatGPT 3.5, beneficiaries can now capture an image of an official letter or copy it within the Cloud app to obtain a coherent, single-click response. Although this flow resembles what vulnerability-bounty platforms faced last year when they were flooded with low-quality automated reports that depleted inspection resources, the researcher notes a fundamental difference: most new government requests come from individuals with genuine statutory rights, but they previously relinquished them under the pressure of what the regulatory literature calls the administrative burden.

Service institutions find themselves facing a complex financial and operational dilemma, as they must process five times the previous number of requests using the same approved resources and budgets.Schmitz argues that viewing this flood as unwanted mail overlooks a rare opportunity to redesign the public care and service system for the agent era; historically, complex procedures acted as a unofficial barrier that limited actual demand, and with AI removing those barriers, institutions are compelled to reconsider the entire procedural framework rather than merely managing the shortfall amid the surge of accumulated transactions.

For digital transformation heads and unified government platform officials in the Gulf, Egypt and the broader Arab region, this shift demands a comprehensive rethinking of back-end intake architecture. Electronic service portals in the region have relied on simplifying front-end submission interfaces while retaining human review in back-office units, a model that will face intense pressure as reviewers begin to use smart agents to draft municipal objections, complaints and formal appeal memos with high precision. The solution will not lie in scaling up human audit staff, but in investing in automated sorting systems that rely on direct data matching, updating standards for processing written forms, and training teams to distinguish between legitimately automated claims and content generated to mislead or flood official channels.

Recognizing the distinction between a flood of useless data and the facilitation of legitimate rights claims is the first step toward regulating this growing phenomenon. The study conveys Schmitz’s optimism that positive user experiences in retrieving taxes and handling complex transactions demonstrate a beneficial role for digital assistants, making the present moment the appropriate time to rebuild governmental procedures from the ground up, lest citizen-empowering tools become a burden that hampers administrative responsiveness.