Waters: Integrating Anthropic AI Into Laboratory Workflows

Waters Corporation, a global leader in life sciences and diagnostics, wants to connect AI directly to its laboratory instruments. This would help pharmaceutical researchers turn ideas created by AI into real experiments with less manual work.

A key part of this plan is Waters AI Link, a new software feature that brings Anthropic’s Model Context Protocol (MCP) to the waters_connect informatics platform.

This connection lets scientists use everyday language to control Waters instruments and manage lab work through AI models and agents, including Anthropic’s Claude and Claude Science.

AI Link can help with tasks such as creating methods, running experiments to collect data, and processing results. It can also coordinate connected equipment, including systems that prepare samples.

Waters’ announcement is another example of a wider change across the industry. More research labs are using AI and automation, bringing computer-based research and real-world experiments closer together.

Danaher has announced that it is developing an AI-powered autonomous lab and has formed a partnership with NVIDIA. Roche is also moving forward with AI-powered research through its own lab-in-the-loop strategy.

Closing the loop

The goal is to connect computer-based research directly with physical experiments. AI models can now generate ideas very quickly, so researchers face growing pressure to test those ideas efficiently.

By automating the steps between planning an experiment, operating instruments and analyzing data, AI-powered workflows help teams test more ideas, reduce the need to pass work manually between people, and speed up research and development.

“We’re not asking anyone to rebuild their lab. We’re bringing AI to the tools they already trust and use. ”

Dave LeithamVice President of Informatics at Waters Corporation

Waters is also joining Anthropic’s Model Hardware Standard (MHS) research programme. This programme aims to make it easier for AI agents to operate physical laboratory equipment safely and independently.

The standard is designed to reduce the time needed to connect instruments to these systems. A process that can take weeks or months could potentially take only hours or minutes.

An important question is how these workflows can be used not just in research but also in regulated pharmaceutical development and quality control.

Waters says its informatics platforms provide a way to move work from waters_connect into its Empower software, which supports regulated laboratory work. This helps maintain the data accuracy, completeness and tracking records needed as molecules move closer to being approved for release.

AI Link is available now through an early-access programme. Waters plans to make it generally available to waters_connect customers in November 2026.

Dave Leitham, Vice President of Informatics at Waters Corporation, spoke with Healthcare Digital about which lab tasks are most likely to benefit from AI, how automated experiments could change pharmaceutical research, and why trust, responsibility and compliance matter as AI enters regulated work.

Which laboratory tasks or workflows do you expect to benefit most from Waters AI Link, and how will scientists and researchers be able to measure the efficiency gains?

DL: The biggest improvements will come from the work around each experiment. This includes building lists of samples, setting up methods, assigning instruments, sending work for processing, and collecting results and reports. Today, these tasks involve a lot of clicking and passing work from one person or system to another.

With AI Link, a scientist explains what they want in everyday language, and the request becomes a specific action in waters_connect software. Newer analysts can also learn the work more quickly, and labs become less dependent on a small number of highly experienced users. Because AI Link can coordinate connected devices, such as sample preparation systems, the different steps can work together instead of operating separately.

To measure the improvement, labs should first record how they work today. They can then compare the time from receiving a sample to reviewing its result, how much their instruments are used, how many manual steps are needed, how often analyses must be repeated, and the cost of each result.

I wouldn't promise the same improvement for every lab, but in development work, we expect timelines to fall from weeks to days.

How could AI-powered, automated laboratory workflows change the speed and scale of pharmaceutical research?

DL: AI can now generate ideas faster than any lab can test them, so the main delay has moved to the lab bench. Every idea produced by a model becomes a real sample that someone must separate and measure.

What excites me is connecting all these steps. AI suggests an experiment, an instrument carries it out, and the result goes straight back to the model. The model then uses that result to plan the next experiment. Each result helps make the next experiment better. Teams waste less time repeating unhelpful steps and can explore many more possibilities.

That's where AI Link comes in. It connects AI agents to data and instruments through waters_connect Software, so scientists don't have to move every step manually from one system to another. As labs adopt this approach, we expect experiments to run more continuously and labs to handle more work without needing to increase staff numbers at the same rate.

The biggest benefit for scientists is time. They spend less time moving work between systems and more time doing what only people can do: deciding what to investigate next and understanding unexpected results.

“The big win for scientists is time. Less of it goes to shuffling work around, and more goes to the part only people can do. ”

Dave LeithamVice President of Informatics at Waters Corporation

What are the biggest challenges in moving AI-directed experiments from research into regulated development and quality control?

DL: It comes down to trust. In research, some variation is part of exploring new ideas. In regulated development and quality control, labs must show exactly what happened, who carried out the work, and that the same process would produce the same result again. This requires reliable data, clear records of every step, proper validation and clear responsibility.

Regulators are still working out what they expect from AI, too. So labs need an approach that they can document and explain clearly if challenged.

That's why the way AI is introduced matters as much as what it can do. With AI Link, we use the Model Context Protocol, an open standard, to build Waters' knowledge of its software and data into the system. This means customers don't have to worry about handling all those details themselves.

Scientists can simply ask for what they need because the relevant expertise is already built in. This gives labs the flexibility and ability to change how they work and use that expertise across their wider workflows. Since everything runs through waters_connect Software, users can still only carry out actions they already have permission to perform.

Different customers will also need different solutions. Many research and development customers will use waters_connect, and AI Link is designed for them. Customers working in quality assurance and quality control (QA/QC), where Empower is a market leader, have their own rules for regulated work. We are also developing AI solutions for them within Empower.

We're not asking anyone to rebuild their lab. We're bringing AI to the tools they already trust and use.

Waters: key partners

Anthropic

CEO: Dario Amodei | HQ: San Francisco, California

Anthropic is an AI research company that develops the Claude family of AI models. Its work with Waters brings the Model Context Protocol (MCP) to laboratory software and includes Waters in Anthropic's Model Hardware Standard research programme. The partnership aims to help AI agents work directly with analytical instruments, making experiments more automated and research workflows more connected.

LabWare

CEO: Vance Kershner | HQ: Wilmington, Delaware

LabWare develops laboratory information management systems (LIMS) and electronic laboratory notebooks (ELN). It has worked with Waters for many years to connect its laboratory software with Waters' Empower Chromatography Data System. This allows labs to transfer sample worklists and analytical results between the two systems. The connection helps reduce manual data entry, protect data integrity and improve workflows in pharmaceutical research and regulated testing.