Anthropic Uses Claude AI to Accelerate Drug Discovery: Could Artificial Intelligence Help Scientists Develop New Medicines?

Developing a new medicine can take years of research, laboratory experiments, and clinical trials.

Scientists must identify what causes a disease, find possible treatments, test whether they work, and carefully examine their safety.

Artificial intelligence could help make some of these steps faster.

On October 27, 2026, Anthropic, the company behind Claude AI, published early findings from its work with biotechnology and pharmaceutical companies.

The company says Claude is helping researchers investigate drug candidates, analyse scientific information, and organise complicated research activities.

Anthropic has been working with six organisations developing new medicines, including Bristol Myers Squibb and BigHat Biosciences.

The company has also started its own early-stage drug discovery research.

But can AI really help scientists discover new medicines, and how close are these experiments to helping patients?

What Is AI-Powered Drug Discovery?

Drug discovery is the process of identifying and developing substances that may prevent or treat diseases.

Traditional drug development usually involves several stages.

Researchers begin by studying a disease and identifying a biological process that could be targeted by a medicine.

They then investigate possible substances that might affect that process.

Promising candidates undergo laboratory testing and further research before they can progress towards clinical trials involving people.

Artificial intelligence can assist scientists by analysing large amounts of information, identifying patterns, and suggesting possible directions for experiments.

For example, an AI system might help researchers compare different proteins or identify changes that could improve a potential medicine.

However, suggestions produced by AI still need to be tested.

A computer-generated prediction does not prove that a treatment will work.

What Is Anthropic Doing With Claude?

Anthropic says it is exploring how Claude can assist with different parts of drug development.

The company has established research collaborations with six organisations:

• BigHat Biosciences
• Bristol Myers Squibb
• Divergence Labs
• Enveda
• Generate
• The Column Group

These organisations are investigating different applications of AI in biotechnology.

Some are exploring how Claude can help improve molecules.

Others are using it to review research information, prepare documentation, or assist with clinical trial planning.

Anthropic is also studying how AI could reduce the administrative work involved in coordinating scientific research.

How Is Claude Helping Scientists Develop Medicines?

1. Improving Potential Cancer Treatments

One research collaboration involves Generate.

The company develops potential medicines using machine learning and biotechnology.

Working with Claude, its researchers investigated a component of CAR T-cell therapy.

CAR T-cell therapy is a type of cancer treatment involving a patient’s immune cells.

Scientists modify certain immune cells to help them recognise and attack cancer cells.

The research team used Claude to help redesign part of the engineered receptor that tells the immune cell when to attack.

According to Anthropic, the team identified a promising design with a signalling component approximately 70% shorter than the standard version.

In laboratory tests involving human T-cells, the redesigned component produced encouraging results.

However, these findings are still at an early research stage.

They do not establish that the redesigned therapy is safe or effective for treating cancer patients.

Further laboratory research and clinical studies would be required.

2. Improving Antibody-Based Drug Candidates

Anthropic also worked with BigHat Biosciences to improve potential antibody-based medicines.

Antibodies are proteins that can recognise and attach to particular targets, such as proteins associated with disease.

Scientists can design antibody-based treatments to interact with specific biological targets.

However, a useful antibody must satisfy several requirements.

It must attach effectively to its target, remain stable, and avoid unwanted clumping.

BigHat’s scientists used Claude alongside laboratory systems to explore ways of improving antibody candidates.

The researchers produced and tested approximately 1,010 molecular variants over about 10 weeks.

According to Anthropic, the selected candidates showed improvements in binding, stability, and resistance to clumping.

Some experimental variants achieved binding improvements exceeding tenfold compared with their starting designs.

These results are promising, but better laboratory properties do not automatically mean the molecules will become successful medicines.

3. Helping Researchers Plan Clinical Trials

Another collaboration involves Bristol Myers Squibb, a global pharmaceutical company.

Clinical trials are studies involving human participants that help researchers evaluate the safety and effectiveness of treatments.

Planning them can require extensive analysis of scientific evidence, possible risks, and previous research.

Anthropic says Bristol Myers Squibb is using Claude to help review clinical development plans and assess possible outcomes.

The company reports that Claude generated an initial analysis that would otherwise have taken considerable manual effort.

However, AI-generated assessments cannot determine with certainty whether a clinical trial will succeed.

Researchers and medical professionals remain responsible for evaluating the findings and making decisions.

4. Supporting Drug Development Documentation

Developing medicines also involves preparing extensive documentation for regulators.

Researchers must maintain accurate records of experiments, safety information, manufacturing processes, and clinical development plans.

Anthropic says Enveda has been exploring Claude-assisted workflows for some of these documentation tasks.

The goal is to help scientific teams prepare and review complex information more efficiently.

However, these documents require careful expert review.

Incorrect information in pharmaceutical documentation can have serious consequences.

Anthropic Is Also Developing Treatments for Rare Diseases

Beyond its external partnerships, Anthropic has begun working on early-stage drug research internally.

One example involves Mithril Medicine, a small research company acquired by Anthropic.

Mithril has been investigating potential treatments for two rare forms of muscular dystrophy.

Muscular dystrophy refers to a group of inherited conditions that progressively weaken muscles.

Some forms begin in early childhood and can be extremely serious.

Anthropic says Mithril’s founder used Claude to help coordinate research activities involving 11 external laboratories.

Claude assisted with organising documents, analysing research information, and managing communication.

The research programme produced encouraging early results in mice.

In one study, treated mice showed improvements in muscle strength compared with untreated mice.

However, these were animal experiments.

The findings do not prove that the treatment will work safely or effectively in humans.

Anthropic says it hopes to move a promising candidate into human clinical trials as early as late 2028, depending on the results of further research.

This is a target, not a confirmed clinical trial start date.

Can AI Really Discover New Medicines?

AI can already help scientists explore possible drug candidates and analyse large datasets.

For example, it can assist researchers in identifying patterns in molecular information and comparing results from many experiments.

It may also reduce the time needed for some administrative and analytical tasks.

However, biology is complicated.

A substance that produces encouraging results in computer simulations may fail during laboratory experiments.

A treatment that works in animals may not work in humans.

Even medicines that perform well during early human trials may later prove ineffective or unsafe.

For these reasons, AI cannot eliminate the need for laboratory testing, clinical trials, and regulatory oversight.

Will AI Replace Medical Researchers?

Anthropic’s research does not demonstrate that AI can independently replace teams of scientists.

Instead, the company is exploring how AI tools can support researchers in scientific and operational work.

For example, Claude may help scientists compare research papers, analyse experimental data, or identify questions worth investigating.

However, researchers still need to decide which experiments are appropriate, conduct studies, interpret results, and evaluate risks.

Human expertise is particularly important when making decisions that could affect patient safety.

What Could AI Drug Discovery Mean for Africa?

African countries face significant healthcare challenges, including infectious diseases, limited research funding, and unequal access to medicines.

AI-assisted research could potentially help scientists investigate diseases affecting African populations.

For example, researchers might use AI tools to analyse scientific literature about malaria or support work involving diseases that receive comparatively limited commercial research funding.

However, the Anthropic announcement does not confirm that its collaborations have developed treatments for these diseases.

The potential benefits would also depend on access to research funding, laboratory facilities, skilled professionals, and appropriate medical regulation.

AI tools alone cannot solve healthcare inequality.

QueryFun Take: AI Could Become an Important Research Assistant

Anthropic’s latest research shows that artificial intelligence is becoming useful in scientific activities that go beyond generating text and answering questions.

AI tools are helping researchers analyse information, design experiments, and investigate possible treatments.

However, the reported progress is mainly early-stage research.

It would be misleading to suggest that Claude has already discovered proven cures for cancer or muscular dystrophy.

For students interested in AI, medicine, and data science, the development demonstrates how skills in biology, programming, mathematics, and data analysis can work together.

Future scientific breakthroughs may increasingly involve collaboration between researchers and AI systems.

Frequently Asked Questions

Is Anthropic using Claude to discover medicines?

Yes. Anthropic is collaborating with biotechnology and pharmaceutical organisations to explore how Claude can assist with drug development.

Has Claude discovered a cure for cancer?

No. The company has reported early research results involving potential treatments, but these are not established cures.

Which companies are working with Anthropic?

The six named research collaborators include BigHat Biosciences, Bristol Myers Squibb, Divergence Labs, Enveda, Generate, and The Column Group.

Can AI replace laboratory experiments?

No. AI can assist with predictions and analysis, but experimental testing remains necessary.

Are the treatments being tested on humans?

Most of the highlighted research remains preclinical. Some collaborations also involve planning or analysing clinical development programmes.

Could AI make medicines cheaper?

AI could potentially reduce some research and development costs, but this does not automatically translate into lower medicine prices.

When could Anthropic’s muscular dystrophy research reach human trials?

Anthropic has identified late 2028 as a possible target for one programme, depending on successful further development.

Primary source: Anthropic: Early Progress on Accelerating Drug Development With Claude