TLDR: The UK is increasingly leveraging AI to enhance the selection process for research funding, particularly emphasizing projects with high commercialization potential. Imperial College London’s “Climate Solutions Catalyst” program exemplifies this, using an AI trained on over 10,000 research abstracts to identify promising green chemistry projects. While AI acts as a powerful filter, human experts still make the final decisions. This approach aims to democratize funding access and uncover overlooked innovations, though concerns about potential biases and data confidentiality persist.
The United Kingdom is at the forefront of integrating Artificial Intelligence (AI) into its research funding landscape, with a significant drive towards identifying projects that demonstrate strong commercialization potential. This strategic shift aims to streamline the traditionally complex and expertise-intensive process of allocating research grants and investments in tech startups.
An international journal, “Science,” recently highlighted the “Climate Solutions Catalyst (CSC)” program at Imperial College London as a prime example of this innovative approach. Launched last year with philanthropic support, the CSC program specifically targets climate-related research nearing commercial viability, offering seed funding without traditional obligations like equity or patent rights. The initiative is powered by an AI developed by researcher César Quilodrán Casas, which was trained on extensive research and industry application data, akin to conversational AI models like ChatGPT. This training enabled the AI to pinpoint green chemistry research with high industrial applicability.
In a notable application, the research team fed over 10,000 abstracts from UK researchers dating back to 2010 into the AI system. The AI initially filtered this vast dataset down to 160 papers deemed to have high commercialization potential. Following this, a joint review by both experts and non-experts further narrowed the selection to 50 papers, from which simple proposals were solicited. Ultimately, three projects received seed funding, including a technology by Professor Joanna Sadler at the University of Edinburgh, which focuses on decomposing disposable plastic tableware into acetone using microorganisms. Professor Sadler expressed her surprise and gratitude upon receiving “unconditional funding of £35,000 (approximately 6.6 million won)” via email.
Christopher Waite, Chief of Scientific Innovation at CSC, emphasized the program’s intent to “uncover groundbreaking discoveries that generally go unnoticed and provide researchers with the tools to bring their findings to market.” This model is seen as a way to level the playing field, as Professor Dashun Wang of Northwestern University suggests that “AI-based pre-analysis could provide a fairer approach for researchers with less commercialization experience or weaker networks,” potentially mitigating biases observed in traditional patent reviews where male professors often have an advantage.
However, the role of AI is not absolute. George Richardson, Head of Data Science at the UK innovation foundation Nesta, clarified that AI functions primarily as a “large filter” to manage the sheer volume of research, with human judgment remaining crucial for final decisions. This hybrid approach is considered effective in proactively identifying suitable candidates for specific challenges.
Despite the perceived benefits, concerns regarding AI in funding reviews persist. Professor Ramana Nanda of Imperial College London cautioned that “AI-based decision-making in venture capital could amplify the bias toward startups that resemble companies that have previously succeeded,” potentially stifling truly novel innovations that deviate from established success patterns. Furthermore, there are significant worries about data confidentiality, with fears that submitted research could inadvertently leak into AI models used for training. Consequently, institutions like the U.S. National Institutes of Health (NIH) and UK Research and Innovation (UKRI) have implemented guidelines prohibiting or restricting the use of generative AI in proposal reviews. Richardson underscored the need for “more testing… to understand the actual impact of AI tools on outcomes.”
Also Read:
- BBC Pioneers Generative AI for Enhanced Newsroom Efficiency and Accessibility
- South Korea’s Ministry of Foreign Affairs to Launch Generative AI System ‘MOFAI’ in November
This initiative reflects a broader trend, with the U.S. Federation of American Scientists (FAS) also proposing AI integration into subsidy reviews to summarize complex research and promote fairness.
