Anthropic reportedly establishes a wet lab in the San Francisco Bay Area, using Claude to direct robots in experiments

Amid the uproar over "AI doomerism," Anthropic has been revealed to have set up a biological laboratory in the San Francisco Bay Area to advance drug science research.

According to a Reuters report on September 19 local time, two sources said Anthropic has built a wet lab in the San Francisco Bay Area. Anthropic hopes to use Claude AI to direct robotic units to carry out scientific experiments with limited human intervention. Anthropic said the goal is to develop treatments for rare diseases, and that its biological research has gone beyond the scope of computer-simulation evaluation, but it will not engage in clinical trials, avoiding competition with pharmaceutical manufacturers.

A so-called wet lab is one in which research is conducted using molecular, cellular and physiological experimental methods in a laboratory. Eric Kauderer-Abrams, head of life sciences at Anthropic, confirmed the existence of the wet lab.

"To do biology, the ultimate test is still real lab work," he said. "We are absolutely doing that today, and I would describe our approach as typical of most biotech companies: some of it in our own facilities, some of it with external partners."

However, a spokesperson later clarified that Anthropic's lab is not specifically for drug discovery, but declined to elaborate. In addition, the spokesperson said human oversight and participation are crucial to safety.

Over the past two weeks, some Anthropic researchers have warned that AI could lead to human extinction. Jacob Coxon, a researcher who has worked on large-model pretraining at both OpenAI and Anthropic, wrote a post warning that the two companies "are racing toward self-improving superintelligence, gambling with everyone's lives." Jacob resigned from Anthropic in early September. He said that among those seriously involved in developing these frontier AIs, quite a few privately truly believe that humanity could go extinct before the end of this decade.

Evan Hubinger, Anthropic's alignment lead, publicly expressed support, saying Coxon is right and that the probability of "AI possibly killing all humans" within 10 years is greater than 10%.

Anthropic has also said it has found some cases in which its systems could potentially be used for biological weapons development.

"In our life sciences work, we are always balancing these factors," Kauderer-Abrams said. "We strive to make models better in order to improve and accelerate the new drug development process, while carefully considering deployment methods to reduce potential risks." He said life sciences are the biggest opportunity for AI to benefit the world, and that we are currently in the early stages of using AI to automate laboratory work, a field with the potential to significantly accelerate many different processes.

For Anthropic CEO Dario Amodei, this effort is personal. He recently said in an article that a disease took his father's life, and just a few years later a cure emerged.

Anthropic is accelerating its layout in the life sciences field. In June, at an event held in San Francisco, Anthropic explicitly stated that it would launch a drug R&D project. At the time, Kauderer-Abrams said Anthropic would conduct preclinical research in areas that traditional companies consider financially unattractive.

Anthropic also launched Claude Science software and acquired startup Coefficient Bio for about $400 million. Anthropic confirmed the acquisition, saying it will help build drug development tools, but declined to comment on the transaction price.

Some diseases have long been difficult to conquer because they are too hard to treat, and AI is expected to accelerate R&D for diseases previously considered "untreatable." But Kauderer-Abrams said that partly out of competitive concerns, Anthropic has set boundaries and currently does not conduct clinical trials, instead focusing on needs the industry has not yet addressed. "We do not compete with pharmaceutical and biotech companies whose business model is to bring drugs to market."

(Original title: AI steps into real biological experiments! Anthropic revealed to have set up a "wet lab," how big is the impact?)

5; Nobel laureate Levitt: AI will be applied to every link of drug R&D

6; After graduating from Tsinghua and doing science popularization for 10 years, how do I fight AI

7; OpenAI adjusts security issue disclosure mechanism

8; The debate over whether AI should hit the brakes is raging, are tech giants really serious?

1; FDE From lesson preparation to grading, AI is reshaping the classroom, but there is one thing it cannot do

2; FDE Under the technology carnival, who is bearing the cost of AI education?

3; Paper topic inspiration: "Sustainable urban and rural development" research direction | MDPI Sustainability

4; Paper topic inspiration: Cerebrovascular diseases (I) | MDPI Brain Sciences

5; AI + multi-source data new framework provides a new solution for agricultural remote sensing mapping

6; Paper topic inspiration: "Sustainable development: agricultural and forestry waste treatment and reuse" research direction | MDPI Sustainability

7; LAF From "intelligent teaching assistant" to "discipline engine": how AI is reshaping university classrooms

8; Paper topic inspiration: "Glioma" research direction | MDPI Brain Sciences

1; Western cryosphere accelerating "degradation": glacier melting sounds the alarm for alpine safety

2; The 6th National Agricultural Science and Technology Development Strategy Think Tank Alliance Conference held in Lanzhou

3; The 2026 Nobel Prizes are about to be announced, authoritative experts give you a comprehensive interpretation!

8; Without a paper, he still graduated with a doctorate, becoming the first person in the national medical field

Famous doctoral dissertations 163: Nobel laureate Peter Doherty's doctoral dissertation, University of Edinburgh, 1970

My elite-school complex and elite-school experience

More>>