AI Transforms Laboratory Practices: Insights from New Research

Research involving 150 laboratory professionals from the United States and Europe reveals that many laboratories currently operate in what is termed a ‘passive state.’ In this environment, electronic laboratory notebooks (ELNs) function mainly as digital filing cabinets, failing to significantly influence research outcomes. Andrew Wyatt, Chief Growth Officer at Sapio Sciences, discusses how the transition to more advanced lab practices occurs through the implementation of the AI Lab Notebook (AILN), which integrates intelligent technology to enhance data connection and decision-making.

The prevailing assumption in discussions around AI in life sciences is that laboratories can seamlessly shift from traditional tools to intelligent platforms. However, Wyatt emphasizes that this adoption is more complex, driven by smaller changes and the daily challenges faced by scientists at the bench. When existing ELNs do not facilitate interpretation or planning, scientists adapt their processes. Over time, these adaptations create a maturity model featuring three distinct stages: passive, shadow, and active.

In the passive stage, the ELN primarily serves as a digital filing cabinet. Although experiments are documented, and compliance is ensured, the software does not actively assist in the next steps. Wyatt notes, “Interpretation, planning, and reuse of results occur elsewhere, often through manual spreadsheets or heavy reliance on specialist informatics teams.” This lack of engagement creates a significant barrier to scientific discovery; 65 percent of scientists report repeating experiments due to difficulties in retrieving or reusing results from their current tools. “These labs do not lack talent,” Wyatt explains. “They are constrained by tools designed to capture past activity rather than actively support scientific reasoning.”

The next evolution arises with shadow labs, where scientists seek to overcome these limitations by employing public generative AI tools alongside their ELNs. While these tools enhance local productivity, they can compromise governance and data integrity. “Seventy-seven percent of scientists report using public AI tools for lab work, and nearly half do so through personal accounts outside organizational visibility,” Wyatt reveals. Shadow labs represent an adaptive response to unmet needs, but they are inherently unstable as they often move sensitive scientific reasoning into unvalidated environments.

Active labs take a more innovative approach by embedding intelligence directly into the notebook environment through the AILN. This transition establishes the notebook as a governed co-scientist rather than merely a supplement. “In an active lab, the AI lab notebook helps interpret results, expose patterns, and connect related experiments in context,” Wyatt states. This integration allows for a more cohesive workflow where designs translate into actionable tasks, and data flows seamlessly between instruments, analysis, and the experimental record.

While active labs do not indicate complete automation, they signify a closer relationship between data, analysis, and action, maintaining human oversight. A crucial aspect of the AILN is that it offers agency rather than merely generating text independently. Scientists can request the notebook to analyze results, compare experiments, or prepare next steps, enabling the system to act on these commands within approved processes. Wyatt points out, “Active labs succeed not by automating decisions, but by reducing friction between observation and understanding.”

Trust remains a central concern in the adoption of AI technologies. Research indicates that 81 percent of scientists will only rely on AI suggestions if they can review the underlying science and evidence. Wyatt cautions that the maturity model is not a diagnostic tool but a practical roadmap for navigating how AI is already entering laboratories. He advises organizations in a passive state to focus on improving data findability, reuse, and interpretation. Enhancing access to existing records can reduce delays and prepare the groundwork for more advanced capabilities.

For labs operating within a shadow state, Wyatt suggests that realism, rather than restriction, is necessary. The transition to the active stage requires strengthening the foundations that link data generation, analysis, and execution into a continuous lab-in-the-loop workflow. As AI models become increasingly capable, laboratories that effectively treat the notebook as a system of reasoning rather than a passive archive are more likely to succeed.

Dr. Tim Sandle, serving as the Editor-at-Large for science news at Digital Journal, contributes expertise in science, technology, environmental issues, business, and health journalism. His background as a practicing microbiologist complements his interests in history, politics, and current affairs.