How AI is transforming in vivo research

AI first appeared on the agenda at the AALAS and FELASA conferences over the past few years, but it always remained a conversation of potential. Today, that has changed. In vivo teams are no longer simply curious about AI; they are actively using it every day. At Benchling, the question heard each day from companies, vivariums, and CROs has moved from "Could AI help us?" to "How can we apply it for maximum value?"

Image Credit:Shutterstock.com/Alcato 

In vivo research methods are advancing rapidly to match the pace of emerging therapeutic modalities. Digitization has increased this transition, allowing teams to gather more data in less time. This data is powering a new generation of intelligent tools. However, fuel has value only when it is clean.

The most important lesson from the past two years is straightforward. AI performs only as well as the structured, connected data that supports it. The teams achieving meaningful results are those that prioritize their data foundation first and gain the benefits across study design, in-life data capture, and analysis.

Below are three areas where AI is delivering practical day-to-day value for in vivo teams.

1. Turning CRO and legacy data from a bottleneck into an asset

Nearly every in vivo program includes work performed externally, ranging from assays on samples to complete studies conducted by a CRO. That data is returned in various formats, including PDFs, Excel, Word, or PowerPoint files, and someone must manually reorganize it before it can be consolidated, visualized, and compared. The process is time-intensive, error-prone, and quietly locks valuable insights inside files that are not easily searchable.

Benchling initially developed early prototypes to address this specific challenge, which currently represents an area where AI provides highly evident advantages. Benchling AI agents can directly process a CRO data package. Results are automatically connected to the appropriate study, animals, and test articles, eliminating manual re-entry and preserving data lineage between the outsourced results and the internal record.

The same method can be used for legacy data stored in older systems. Rather than leaving years of historical studies isolated, teams can convert that data into a structured, searchable format and use it to refine the parameters for future experiments. This capability transforms in vivo data from an isolated silo into an integrated part of the connected R&D landscape.

2. AI that saves scientists' time on reporting

Approximately half of a scientist's workweek is spent on manual, logistical tasks such as data preparation, management, and reporting instead of working on novel science. Reporting, in particular, remains a persistent burden. It frequently requires compiling a study report during the study or at takedown, summarizing successful outcomes and challenges for managers and colleagues, and reformatting data to meet downstream requirements.

This is precisely the type of work that AI is now handling. When applied to structured study data, AI can produce a high-level study report within minutes, significantly reducing a task that traditionally took an entire afternoon of copying and pasting. It can also perform specialized formatting tasks that previously required costly consulting services or additional competitor software.

One important application is verifying data against regulatory submission requirements such as SEND. AI can now perform the bulk of this work, while the qualified scientist serves as the final reviewer.

After a report has been generated, the same tools can translate it as needed. This represents a significant advantage for global teams that require the same study summary in multiple languages. The consistent pattern across all these applications is that AI manages the repetitive workload, allowing the scientist to concentrate on the critical and specialized judgment that only they can provide.

3. Anomaly detection and statistical analysis as a built-in safety net

Collecting data in a structured, objective manner is challenging, particularly when multiple scientists contribute to the same dataset. Errors can easily occur, such as a tumor volume increasing more than anticipated between measurements or a duplicate entry recorded on the same date. These issues often remain unnoticed until they are identified during review, if they are detected at all.

AI enables teams to verify the data. It can determine whether anomalies exist in a study's tumor volume results or whether a liver sample was collected from every animal. It can also identify issues in completed studies as well as those still underway, allowing problems to be addressed while there is still an opportunity to take corrective action.

The same structured data foundation also supports faster analysis. Because the data resides within a single platform, AI can perform a complete statistical evaluation by checking normality, selecting the appropriate test, and identifying when a reported p-value no longer remains significant after correction. This helps prevent results from being overinterpreted. By applying this level of analytical rigor and making it accessible to every scientist, regardless of statistical expertise, every result remains traceable to its original source data. This safeguards study integrity and the animals - fewer unnoticed errors result in fewer repeated experiments.

The thread that ties it together

None of these three outcomes are really about applying AI on its own. CRO or legacy data ingestion, automated reporting, and anomaly detection all rely on the same foundation - governed, structured, and connected data that intelligent tools can interpret.

The reality is that free-text observations and proprietary, disconnected data models are essentially inaccessible to the AI ecosystem that R&D organizations are developing.

Based on Benchling’s experience, the teams achieving the greatest value from AI are those that view their data foundation as a prerequisite rather than an afterthought. This trend is driving the next generation of in vivo studies.

Benchling In Vivo is specifically designed for this purpose. It records in-life data in a structured format, links it across upstream and downstream teams, and ensures the information is prepared for Benchling AI to convert into meaningful insights. At the same time, it maintains the highest standards of 3Rs compliance and animal welfare.

See it live in Houston at AALAS 2026

The team will be returning to the AALAS National Meeting in Houston, TX between October 25–29, 2026, where Benchling welcomes the chance to continue these discussions face-to-face.

Find Benchling at:

• Catch the talk. Attend the Technical Trade session, "AI-Assisted In Vivo Research: Real-World Applications," scheduled for 1:00 pm on Sunday, October 25.
• See it live. Visit booth 1146 to see live walkthroughs of the AI capabilities highlighted throughout this article, including CRO and legacy data ingestion, AI-assisted reporting, along with anomaly detection and analysis.
• Talk it through. Engage with the creators and daily users of these technologies to evaluate how AI can integrate into the existing in vivo workflows.

About Benchling

Benchling makes biotech research and development faster and more collaborative. Biotechnology has the potential to solve humanity’s most pressing challenges, such as disease, renewable energy, clean water, and hunger. The brightest minds are working on these problems but they are equipped with archaic tools. We aspire to fix this and increase the rate of scientific output with a web-based platform that allows researchers to design and run experiments, analyze data, and share results.

Hundreds of thousands of scientists all around the world use Benchling to do research. Whether they are at the world’s largest companies, the top research universities, or working on a startup in a garage, scientists use Benchling for the same reason: to be empowered, not encumbered, by their tools.

Sponsored Content Policy: News-Medical.net publishes articles and related content that may be derived from sources where we have existing commercial relationships, provided such content adds value to the core editorial ethos of News-Medical.net, which is to educate and inform site visitors interested in medical research, science, medical devices and treatments.