AI-Assisted Proposal Writing and Its Implications for Research Funding Competition and Innovation Advancement

• In Event: AI and the Public Workforce – Skills, Concerns, and Services

Abstract

Background

The rapid diffusion of artificial intelligence-generated content (AIGC) is reshaping competitive funding systems, lowering barriers to proposal development while amplifying pressures on resource-constrained agencies. By reducing the time, effort, and expertise required to draft submissions, AIGC broadens participation in funding competitions, enabling applicants with limited resources to engage more actively. However, this technological shift introduces a paradox: while expanded participation aligns with goals of inclusivity, it risks overburdening review systems, inflating submission volumes, and complicating agencies’ efforts to identify high-impact proposals

Research Method

This study employs an integrated economic and computational approach to analyze AI’s systemic impacts. An economic model grounded in contest theory examines strategic interactions between researchers and funding agencies, simulating how AI-driven reductions in proposal drafting costs influence submission behaviors, competition intensity, and agency evaluation policies. The model explores scenarios where researchers adopt AI tools to lower effort thresholds, balancing the trade-offs between increased participation and risks of quality homogenization.

Complementing this, agent-based computational simulations replicate real-world funding environments using large language models to generate autonomous researcher agents. These agents mimic diverse academic profiles and adapt strategies dynamically. The simulations incorporate peer review biases, agency budget constraints, and evolving policies, enabling granular analysis of how AI adoption shifts systemic outcomes.

Key Findings

The findings indicate that AI assistance significantly lowers the barriers to proposal submission, leading to a substantial increase in the number of proposals. This surge intensifies competition for limited funding resources, potentially disadvantaging researchers who lack access to AI tools or who choose not to use them. The increased volume of submissions can overwhelm funding agencies and reviewers, affecting the thoroughness and fairness of the evaluation process.

While AI can enhance the presentation and clarity of proposals, there is a risk that over-reliance on AI may result in homogenization of content. Proposals may become similar in style and structure, making it challenging for truly innovative ideas to stand out. Researchers might invest less effort in developing novel concepts, relying instead on AI to embellish standard ideas. This could lead to a dilution of the overall quality of proposals, with originality and creativity being overshadowed by polished but less substantive submissions.

The heightened competition and potential quality dilution might reduce the likelihood of funding agencies selecting proposals with significant potential for scientific breakthroughs. Reviewers may experience fatigue due to the increased workload, leading to superficial evaluations or reliance on heuristics that do not favor unconventional or groundbreaking proposals. Consequently, the advancement of science and technology could be hindered if the most promising research does not receive the necessary funding.

Implications

Funding agencies must balance AI’s efficiency gains with safeguards for equity and innovation. Clear guidelines on AI use, such as mandatory disclosure of AI tools, could maintain evaluation integrity. Submission limits or tiered review processes may mitigate volume-driven overload. Agencies should invest in reviewer training and AI-assisted triage systems to manage workloads while preserving attention to breakthrough potential. Dedicated funding streams for high-risk projects and equity-focused resource allocation are critical to counter systemic biases.

Author

• Fei Xiao, The Hong Kong University of Science and Technology (Guangzhou)
  Presenting Author