One of the most transformative technologies in clinical research today is the digital twin.
It marks a paradigm shift in how control arms are designed. AI powered models simulate how a disease would progress without treatment, so each participant in the treatment group can be paired with a personalized "digital twin."
Unlearn AI, a U.S.-based company, is leading the field, with its platform already applied in oncology and neurology trials. Its PROCOVA methodology received the first ever EMA qualification opinion for an AI methodology, and is aligned with the FDA's 2026 guiding principles for AI in drug development.
The technology uses historical clinical and biological data to generate synthetic control groups that closely mirror how patients would have responded without the investigational therapy. This speeds up recruitment by reducing placebo enrollment, shortens timelines, and reduces control arm size, without compromising data quality.
According to published data, control arm size can be reduced by roughly a third, Phase 3 recruitment shortened by up to four months, and full statistical power maintained.
Why it matters. For the developing company: faster, more efficient studies. For patients: greater access to advanced treatments, and a lower chance of being placed in the control group. For the industry: more ethical, data-driven trial design.
What is next? We may soon run trials that enroll only treated participants, while their digital twins generate the comparative data once provided by control groups. In 2026 the use is still largely supplemental and hybrid, but the direction is clear. The future is already here.