New AI agents designed to review and rank research proposals are already being tested by grant-giving bodies, but academics warn that their impact could be catastrophic.
Researchers say that these AI agents, while intended to make the assessment process more efficient and less biased, could ultimately lead to a complete overhaul of the current system, potentially favoring big-budget projects and big-name researchers over smaller, innovative initiatives.
The AI Revolution in Grant Funding
The introduction of AI-powered grant review systems is being touted as a way to streamline the notoriously time-consuming and often grueling process of awarding research funding.
However, experts warn that the use of these AI agents could have unintended consequences, including:
- The prioritization of high-impact, high-risk research, which may lead to a decrease in funding for smaller-scale, more practical projects.
- The increased focus on research with commercial potential, which could have a negative impact on basic science research and potentially stifle innovation.
Additionally, the reliance on AI agents could exacerbate existing biases and inequalities, particularly in fields where research funding is already concentrated among a select few.
The Risk of Exacerbating Existing Biases
Academics point out that the data used to train these AI agents is often sourced from existing research databases, which are likely to reflect the same biases and inequalities that the current system perpetuates.
This could lead to a feedback loop, where AI agents perpetuate existing biases, rather than breaking free from them.
What This Means
The rise of AI agents in grant funding poses a significant threat to the diversity and innovation of research projects.
Researchers and policymakers must consider the potential risks and unintended consequences of relying on these AI agents and develop strategies to mitigate them.
Ultimately, the goal should be to create a more equitable and inclusive research funding system, one that values and supports a wide range of projects and researchers, regardless of their size, scope, or budget.
