AI Model Tests 1,758 CAR T-Cell Binders
CHAI president Brian Anderson outlined AI's role in rheumatology and CAR T-cell design at ACR Convergence.
Updated on Nov. 8, 2026 in Arthritis
Live Poll
Do you trust AI to be used responsibly in healthcare?
Brian Anderson, president and CEO of the Coalition for Health AI (CHAI), delivered the keynote during the opening ceremony of ACR Convergence on November 8, 2026. The talk covered AI in rheumatology and the design of CAR T-cell therapies, including an AI model's assessment of 1,758 binders targeting BCMA, CD-19 and CD-22.
Why it matters
The binders were validated through protein binding assays, T-cell activation assays and in vivo killing experiments, which exposed off-target activity, unwanted signaling and inaccessible epitopes that computer design alone could not identify. The findings suggest AI-assisted drug design still requires laboratory testing before it can be relied upon.
An AI model assessed 1,758 binders targeting B-cell maturation antigen (BCMA), CD-19 and CD-22 for CAR T-cell design. The source does not report how many of these candidates passed experimental testing.
The players
Brian Anderson
Brian Anderson is president and CEO of the Coalition for Health AI, a group focused on responsible AI use in healthcare.
Coalition for Health AI (CHAI)
The Coalition for Health AI is a nonprofit organization that works on standards and governance for artificial intelligence in healthcare.
The details
Investigators tested the binders through protein binding assays, T-cell activation assays and in vivo killing experiments. The testing identified off-target activity, unwanted signaling and inaccessible epitopes in CAR T-cell therapeutics.
Timeline
November 8, 2026: Anderson delivers the keynote at the ACR Convergence opening ceremony in Orlando.
Through 2027: The AI sandbox process used in Utah is expected to be reproduced in states across the country.
Health Landscape
CHAI is participating in Utah's AI sandboxes, where regulators decide how AI may responsibly be used in healthcare. The source projects that this sandbox process will be reproduced across U.S. states through 2027, shaping how next-generation treatments like CAR T-cell therapy are developed.
Patients receiving CAR T-cell therapies may eventually benefit from AI tools that screen candidate molecules, though the source notes that laboratory testing still reveals problems such as off-target activity. Physician and patient input in AI development will shape how these next-generation treatments reach clinics.
The takeaway
AI can rapidly screen large candidate pools for therapies, but experimental validation remains essential before results are trusted. Clinicians and patients who engage with AI development can help ensure these tools reflect real-world treatment needs.
What happens next
State-level AI sandbox processes like Utah's are expected to be reproduced across the country through 2027.
Further reading
Explore more on Arthritis to see how emerging AI tools are being applied to rheumatologic disease research.
Source note: This article includes information reported by Healio.
Live Poll
Do you trust AI to be used responsibly in healthcare?
