Federal service contractors spend considerable effort forecasting workload. Staffing, productivity, facilities, technology, and contract pricing often depend on assumptions about how many transactions the Government or the public will generate during performance.
Artificial intelligence may be making those assumptions increasingly unreliable.
In an October 14 essay, Drowning in Slop, public-policy commentator Sam Freedman raises an important question: what happens when AI makes it dramatically easier for individuals to generate complaints, applications, petitions, and other submissions that government institutions are required to process?
The underlying research suggests that this is no longer merely theoretical.
In Characterizing Agentic Flooding of Government Services, researchers Chris Schmitz, Lewis Hammond, and Alan Chan assembled 84 potential cases across 11 jurisdictions in which AI-assisted activity was reportedly associated with increased demand on government services.
The researchers distinguish between increases in the number of submissions and increases in their complexity. An agency may receive more applications, or it may receive submissions that are substantially longer and more demanding to review.
The researchers caution that their evidence does not establish precisely how much of each increase was caused by AI. Nevertheless, the pattern identifies an emerging operational risk.
For federal contractors, the connection is direct.
Consider a contractor operating a government contact center, benefits-processing operation, administrative support function, public inquiry system, or records-review service.
The contractor priced the work based on expected transaction volumes, average processing times, labor productivity, and service-level requirements.
Then AI makes it much easier for individuals to submit additional requests or produce more complex supporting documentation.
The Government's workload increases.
The contractor's revenue may not.
Under a firm-fixed-price contract, that difference can become a significant margin problem.
The contractor remains responsible for the agreed scope and performance standards even if actual volumes exceed its assumptions. An unexpected increase in demand does not, by itself, establish entitlement to an equitable adjustment.
That is why the contractual architecture matters.
Contractors should distinguish between baseline workload risk and changes in Government requirements. A contract that explicitly places broad volume uncertainty on the contractor presents a different entitlement analysis from one in which the Government subsequently expands the services required or changes the processing standards.
Pricing should therefore account for more than historical transaction counts.
Companies should examine whether AI could increase demand, average submission length, verification complexity, exception handling, quality-control requirements, and the amount of human review required.
Where appropriate, proposed pricing structures might include workload bands, transaction-based compensation, surge CLINs, agreed productivity assumptions, volume thresholds, or defined procedures for repricing materially changed requirements.
Contractors should also measure the difference between increased volume and increased complexity. Processing twice as many simple inquiries does not necessarily cost the same as processing the original volume of submissions that are twice as difficult.
There is an important public-policy distinction as well.
AI may improve access to government services by helping legitimate users navigate difficult procedures. Higher demand is not inherently abusive or undesirable.
Contractors should therefore avoid treating every AI-assisted submission as suspicious or recommending barriers that unnecessarily exclude legitimate users.
The better response is to design service systems—and contracts—that can absorb predictable changes in demand while preserving appropriate access and quality.
The larger lesson is familiar in federal contracting.
When technology changes the economics of generating work, contracts based on historical workload assumptions may become financially obsolete.
AI will not merely transform how contractors perform government services.
It may transform how much service the public demands.
Recommended FedContractPros Product: GovCon BoE Blueprint™
The GovCon BoE Blueprint™ can help contractors document workload baselines, labor productivity, transaction volumes, staffing assumptions, surge capacity, indirect costs, and pricing sensitivities when preparing fixed-price service proposals or evaluating contract volume risk.
Disclaimer:
This article is provided for general informational and educational purposes only and does not constitute legal, pricing, financial, artificial-intelligence, or government-contracting advice. The cited research identifies potential cases of AI-related increases in administrative demand but does not establish that AI caused every reported increase. Entitlement to a contract adjustment depends on the particular contract type, allocation of workload risk, Changes clause, applicable requirements, notice, causation, and supporting evidence. Contractors should evaluate these matters before adopting pricing assumptions or pursuing requests for equitable adjustment.