Anthropic's Claude AI Generates 35,000 Antibody Designs for Allergy Treatment

AI model rapidly creates potential drug candidates, accelerating discovery and development processes.

The word "ANTHROPIC" is displayed in white capital letters on a black smartphone screen.
Anthropic's Claude model rapidly generates antibody designs for a potential food allergy treatment.
Key Points

Anthropic's Claude model generated approximately 35,000 antibody designs in about a month for a potential food allergy treatment, aiming to accelerate drug discovery, manufacturing, and clinical development processes.

  • Anthropic's Claude model generated 35,000 antibody designs for a food allergy treatment.
  • The AI collaborated with Divergence Labs on a treatment to remove IgE from immune cells.
  • Three designs performed better than standard medication at dislodging IgE.
  • Claude helped Generate Biomedicines develop a 70 per cent shorter CAR T-cell therapy instruction.
  • At Enveda, Claude reviewed 1,061 pages, reducing human review time by 89 per cent.
  • Anthropic is partnering with Bristol Myers Squibb to support clinical trial planning.

Artificial intelligence company Anthropic said its Claude model generated around 35,000 antibody designs in about a month for a potential food allergy treatment, as it expands efforts to use AI across drug discovery, manufacturing and clinical development.

In an update on its life sciences research, Anthropic said its work with biotechnology company Divergence Labs produced early results for a proposed treatment designed to remove immunoglobulin E (IgE), a molecule involved in allergic reactions, from immune cells. The approach aims to go beyond standard medications that neutralise IgE circulating in the blood.

Divergence Labs identified three designs that could bind to IgE and 20 that could dislodge it from immune cells. Three of the 20 designs appeared to perform better than the standard medication at dislodging IgE, Anthropic said. The next step is to combine the two functions into a potential drug.

The company said it had partnered with six organisations developing medicines and had also started its own drug discovery programmes, including research into rare childhood forms of muscular dystrophy.

“We plan to scale this work up substantially from here,” Anthropic said, adding that its strategy was to involve AI in multiple stages of drug development rather than focusing only on designing molecules.

In another project, Claude helped Generate Biomedicines develop a shorter version of genetic instructions used in CAR T-cell therapy, which is being explored for treating solid tumours. One design was around 70 per cent shorter than the standard version and, in laboratory tests, enabled immune cells to produce roughly 10 times as many receptors and kill tumour cells more efficiently, Anthropic said.

At Enveda, Claude reviewed 1,061 pages of manufacturing documentation, identifying all 11 issues found by human reviewers and flagging seven additional issues. The approach reduced human review time by 89 per cent, from more than 25 hours to less than three hours.

Anthropic also said its work with Bristol Myers Squibb was examining how Claude could support clinical trial planning and assess the likelihood of trials succeeding. The companies plan to expand this work in 2027.

However, Anthropic cautioned that the research remains at an early stage and is mostly preclinical, meaning the potential medicines have generally not yet reached human testing.

The company said it plans to work with more biotechnology organisations and integrate Claude further into drug development as the programmes advance.

Published: Wednesday, 28 October 2026 | 1:40 PM IST

Share this story

WhatsAppFacebookXTelegram