Artificial intelligence is transforming research, development, and innovation (R&D&I) from an experimental tool into a core driver of productivity, creativity, and breakthrough discoveries. Mark Venables examines Arthur D Little’s Eureka! On Steroids report, which explores AI’s potential, the challenges of adoption, and the strategic paths organisations can take to stay ahead.
Artificial intelligence has moved from an experimental tool to a core enabler of research, development, and innovation (R&D&I). The latest report by Arthur D Little, Eureka! On Steroids, sets out how AI is reshaping discovery processes, accelerating decision-making, and unlocking creativity across industries. Companies at the forefront, such as Google DeepMind and L’Oréal, are already reporting exponential gains in productivity.
Yet, AI’s promise comes with challenges. Data silos, talent shortages, and the complexity of integrating AI into established R&D&I workflows remain critical barriers. The report outlines six potential future scenarios, offering organisations a roadmap to harness AI effectively while mitigating risks. Leaders who invest in AI strategies today will define the future of innovation; those who delay may struggle to keep pace.
A moment of transformation
Dr Albert Meige, Director of Blue Shift at Arthur D Little, believes the timing of the report is critical. “There are several reasons, the first is the general context,” he says. “Over the past few years, significant developments have demonstrated AI’s growing role in scientific research. In December 2023, an article in Nature reported that a group of researchers had solved a previously unsolved mathematical problem by collaborating with AI, specifically a large language model. This milestone went largely unnoticed in mainstream media, but it represents a fundamental shift in how AI contributes to scientific creativity.”
He points to a breakthrough from June 2024, when University of Pennsylvania researchers used AI to identify nearly a million new antibiotics, some with promising properties against antibiotic-resistant bacteria. AI-driven discoveries such as these highlight the technology’s potential to tackle previously insurmountable challenges. “Beyond these milestones, many of our clients at Arthur D Little were keen to explore AI’s impact on R&D&I,” he adds. “As a global firm working with major organisations on innovation, we saw the opportunity to bring clarity to this fast-moving landscape.”
AI’s role in redefining research
The report identifies three core ways in which AI is transforming R&D&I: increasing productivity, solving previously unsolvable problems, and improving decision-making. “AI is boosting productivity by automating administrative and repetitive research tasks,” Meige explains. “Literature reviews, report writing, and bibliography compilation can now be streamlined, allowing researchers to focus on core scientific challenges.”
But AI is more than a tool for efficiency. It is also a collaborator in complex problem-solving. “Certain research problems were historically beyond human capability due to their complexity or the sheer volume of data involved,” he continues. “AI provides researchers with new ways to tackle these problems, helping to generate insights and creative solutions that were previously out of reach.”
Zoe Huczok, Project Leader at Blue Shift, highlights AI’s impact on high-stakes industries such as healthcare, energy, and technology. “AI is transforming drug discovery through methods like target-based molecule identification,” she explains. “In clinical trials, AI is improving efficiency by optimising patient group assignments. In the energy sector, AI has long been used for grid balancing and predictive maintenance. More recently, DeepMind demonstrated AI’s role in nuclear fusion by applying reinforcement learning to plasma control.”
Beyond these applications, AI is also making inroads into materials science, where machine learning models predict new materials with desirable properties faster than traditional laboratory experimentation. In manufacturing, AI is automating quality control, detecting microscopic defects that the human eye cannot see, and optimising production workflows for greater efficiency. It is also revolutionising logistics by improving supply chain forecasting, reducing waste, and increasing responsiveness to demand fluctuations. In agriculture, AI is enabling precision farming, analysing soil conditions, predicting crop yields, and automating harvesting processes, which could lead to more sustainable food production.
AI is also emerging as a game-changer in climate research, allowing scientists to simulate environmental changes, model climate patterns, and develop advanced mitigation strategies. The technology is increasingly being used to enhance carbon capture methodologies, improve waste management systems, and drive innovations in sustainable materials.
Overcoming barriers to adoption
While the potential of AI in R&D&I is undeniable, barriers to widespread adoption remain. The report identifies several challenges, including data quality issues, infrastructure limitations, and talent shortages.
“Many organisations have decades of data, but much of it is unstructured, inconsistent, or locked in silos. Without proper data governance, AI models cannot deliver their full value,” Meige says.
The report emphasises three key strategies to address this:
- Improving data quality – AI models are only as good as the data they are trained on. Organisations must invest in cleaning and structuring their data for maximum effectiveness.
- Encouraging collaborative data access – AI thrives on diverse data sources. Breaking down silos and fostering partnerships between business units, suppliers, and academia enhances AI’s impact.
- Leveraging proprietary data – As AI models become more widely available, unique datasets will provide organisations with competitive advantage.
Agile methodologies are also crucial in AI implementation. “AI models require continuous monitoring and adaptation,” Huczok explains. “Traditional software development methods are not always sufficient. Agile approaches allow companies to iterate quickly, ensuring AI remains relevant and effective over time.”
Moreover, successful AI adoption requires cross-disciplinary collaboration. Scientists and engineers must work closely with AI specialists to ensure that models are designed with domain-specific knowledge. Companies must also invest in upskilling their workforce to integrate AI-driven processes effectively.
The future of AI in R&D&I
The Eureka! On Steroids report outlines six possible future scenarios for AI in R&D&I, determined by factors such as performance, trust, and affordability. These range from AI becoming an essential force in every aspect of research to it being limited to specific low-risk applications.
Each scenario presents distinct implications for organisations:
- Blockbuster – AI is deeply integrated across all R&D&I functions, driving an unprecedented wave of innovation. Breakthroughs in materials science, quantum computing, and life sciences accelerate at a pace never seen before.
- Crowd Pleaser – AI adoption is widespread but limited to operational efficiency improvements rather than groundbreaking discoveries. Research becomes more efficient but does not fundamentally change the innovation landscape.
- Crown Jewel – AI’s benefits are concentrated in a few well-funded organisations, exacerbating the gap between research leaders and laggards. Those with the resources to develop proprietary AI models achieve significant advantages.
- Problem Child – AI’s adoption is hindered by ethical concerns, regulatory constraints, and public mistrust. While AI tools exist, their use is constrained by legal and social limitations.
- Best-Kept Secret – AI models achieve high levels of performance, but access remains restricted due to proprietary concerns and security risks. Only a select few industries benefit, limiting widespread AI-driven progress.
- Cheap & Nasty – AI is relegated to low-value, untrusted applications due to poor data quality and lack of regulation, leading to limited adoption outside of specific niche cases.
Each of these scenarios forces organisations to consider their strategic approach to AI adoption. Companies that embrace AI holistically and invest in governance, infrastructure, and talent will be best positioned to succeed regardless of how the future unfolds.
Seizing the AI advantage
For organisations still hesitant about AI’s return on investment, Meige offers clear advice. “AI is already delivering tangible benefits across industries,” he says. “Companies that act now to secure data access, build AI capabilities, and implement governance frameworks will be in the strongest position to capitalise on AI’s potential. Those who delay will find it increasingly difficult to catch up.
“AI is no longer an experimental add-on to the innovation process. It is rapidly becoming an essential component of R&D&I, enabling organisations to push the boundaries of discovery, optimise decision-making, and unlock new levels of creativity. The companies that move decisively today will be the ones shaping the future of innovation.”
