Anthropic is launching its own drug discovery programs for rare diseases using Claude superintelligence

The news landed like a thunderclap in a San Francisco conference room on June 30, 2026. Anthropic, the artificial intelligence company behind the cutting-edge Claude model, announced it will launch its own preclinical drug-discovery programs targeting neglected diseases, including rare conditions, while simultaneously unveiling Claude Science, an AI workbench built for researchers and drug-makers.

This development could become a direct challenge to the pharmaceutical establishment, a sector that has spent decades perfecting the art of maximizing profit margins while leaving millions of patients with rare and overlooked conditions to suffer in silence. If we’re betting on humanity’s best intentions, Anthropic is in a position to unleash super-intelligence that could upend decades of corporate greed and Big Pharma’s exploitation of human patients. But the super-intelligence could go both ways and be leveraged by Big Pharma to continue making customers for life. The deeper question remains: Will this super-intelligence be used to genuinely heal, or will it become the most sophisticated tool yet for manufacturing lifelong customers?

Key points:

• Anthropic will run its own preclinical drug programs for neglected and rare diseases, targeting conditions that are ignored for economic reasons.
• The company launched Claude Science, an AI workbench for researchers, on June 30, 2026, at a San Francisco event.
• Eric Kauderer-Abrams, Anthropic’s head of life sciences, stated the company needs to “live it along with all of you” to build the right tools.
• Rare diseases offer clearer biological targets, often stemming from single damaged genes, making them more amenable to AI-driven solutions.
• Anthropic acquired Coefficient Bio for $400 million and placed Novartis CEO Vas Narasimhan on its board, signaling deep industry entanglement.
• The dual-agent approach used in tools like Cursor Code demonstrates AI’s growing capacity for complex, multi-step tasks like drug discovery.

The hard truth about why your disease gets ignored

To understand what Anthropic is really doing, you must first understand the brutal economics that dictate which diseases get researched and which get abandoned. Major pharmaceutical companies operate on a simple calculus. Developing a single drug can cost anywhere from $1 billion to $2.6 billion when factoring in the cost of failed trials. The process takes ten to fifteen years. And even then, the Food and Drug Administration approves only about ten percent of drugs that enter human trials. For a company like Pfizer or Merck, investing that kind of money into a condition that affects 10,000 people worldwide is financial suicide. The math simply does not work.

This is why thousands of rare diseases have no approved treatments at all. According to the National Institutes of Health, there are more than 7,000 known rare diseases, and approximately 95 percent of them lack any FDA-approved therapy. Patients are told to manage symptoms, to hope, to wait. Behind closed doors, executives admit the truth. The return on investment is too low. The patient populations are too small. The Wall Street analysts would revolt.

Anthropic’s Jonah Cool, the head of life sciences partnerships and deployment, put it bluntly when speaking to STAT. “These are areas that normal drug development economics don’t incentivize or favor.” He added, “The idea here is that the biology is often clear; the economics, if you’re trying to run a drug development business, are challenging.” People are dying, suffering, and deteriorating because the profit motive has failed them. Utilizing super-intelligence, drug researchers could find solutions that don’t depend on these profit motives.

The double-edged sword of super-intelligence

Here is where the story gets both hopeful and deeply troubling. Anthropic’s Claude, like all frontier AI models, can process and analyze biological data at a scale no human team could match. The company’s Claude Science workbench is pre-configured for genomics, single-cell analysis, proteomics, and cheminformatics, backed by more than 60 scientific databases. The company claims every result is reproducible and traced to its code. This is the same kind of architecture that powers tools like Cursor Code, where one agent plans the project and another handles execution, automatically summarizing context and starting fresh when it reaches the end of its processing window. Apply that dual-agent capability to biology, and you have a system that could theoretically identify molecular targets, design candidate compounds, run virtual simulations, and propose delivery mechanisms all within a fraction of the time human researchers would need.

But who controls the intelligence? That is the question that should keep every patient awake at night. A super-intelligence capable of designing novel drugs is also a superintelligence capable of designing novel dependencies. Big Pharma has a well-documented history of taking simple health problems and medicalizing them into chronic conditions requiring lifelong pharmaceutical management. High blood pressure, high cholesterol, acid reflux, anxiety, depression. All natural human variations or responses to environmental stressors. All transformed into profit centers.

The same technology that could design a phytochemical delivery system to activate your body’s own genetic repair mechanisms could also design a molecule that creates a brand new disease classification, complete with a branded drug that must be taken forever.

The same AI that could match your unique microbiome with targeted nutritive compounds could also engineer a dependence that leaves you no alternative but the company’s patented product.

Reason for hope and skepticism

Anthropic’s acquisition of Coefficient Bio for approximately $400 million and the placement of Novartis CEO Vas Narasimhan on its board should give every discerning observer pause. Novartis is one of the largest pharmaceutical companies in the world. It has a revenue stream built on selling patented drugs at monopoly prices. It has every incentive to use whatever tools are available, including super-intelligence, to maintain and expand that model.

Eric Kauderer-Abrams, Anthropic’s head of life sciences, told the San Francisco audience, “We believe in the power of tight feedback loops, and there’s no substitute for having our own experiences alongside you all in the trenches trying to develop drugs.” That sounds noble. But the trenches are dug by whoever controls the shovel. If the feedback loop connects Claude’s super-intelligence to Big Pharma’s profit incentives, the output will inevitably be more patents, more lifelong prescriptions, more conditions that require chemical management rather than genuine healing.

The potential for good is immense. A true benevolent application of this technology would focus on optimizing the use of nutritive elements, phytochemicals, and natural compounds that work with the human body rather than against it. It would design delivery systems that target specific genetic on-off switches, activating repair mechanisms and silencing disease pathways without the toxic side effects of synthetic drugs. It would match treatments to individual microbiomes, recognizing that no two humans are biochemically identical. That is the promise.

But the track record of corporate America, and of the pharmaceutical industry specifically, argues strongly for skepticism. When a company like Anthropic says it will pursue diseases that “normal drug development economics don’t incentivize,” it is simultaneously admitting that it has the ability to work outside those economics. The question is whether it will. Or whether, once the technology proves itself on rare diseases, the same super-intelligence will be turned toward engineering the next blockbuster condition that keeps the prescription pads filled for generations to come.

Sources include:

Zmesscience.com

CNBC.com

StatNews.com