Truace
Subscribe

An AI agent matched hematology tumor boards' decisions about 83% of the time in a silent trial

Study title: From Answers to Agents: What Must Change Before Generative and Agentic AI Become Clinical Infrastructure in Hematology

Generative artificial intelligence in hematology is entering a new phase. The dominant question, whether large language models are accurate enough for clinical decision support, is being overtaken by a harder one, as systems shift from answering questions to acting: extracting structured cases, routing them, classifying variants, and grounding recommendations in guidelines and case memory.

Read the study abstract

Generative artificial intelligence in hematology is entering a new phase. The dominant question, whether large language models are accurate enough for clinical decision support, is being overtaken by a harder one, as systems shift from answering questions to acting: extracting structured cases, routing them, classifying variants, and grounding recommendations in guidelines and case memory. In 2026, a hematology agent achieved roughly 83% concordance with tumor-board decisions in a prospective silent trial, with…

Truvace record2 min read1 sourceEvidence: Emerging · 30/100 — Commentary, opinion, or narrative review · 1 independent sourcePeer-reviewedTRV-2026-1168 · cite or check it
From Answers to Agents: What Must Change Before Generative and Agentic AI Become Clinical Infrastructure in Hematology"Medical Laboratory" by ben.dracup, CC BY 2.0.

In brief

In 2026, a hematology agent achieved roughly 83% concordance with tumor-board decisions in a prospective silent trial, with hallucinations in 0.3%, suggesting that for well-structured tasks the binding constraint is shifting from accuracy to governability. For hematologists, this reframes AI adoption as a governance and human-factors program rather than a search for a better model.

Main points

  1. Generative artificial intelligence in hematology is entering a new phase.
  2. The dominant question, whether large language models are accurate enough for clinical decision support, is being overtaken by a harder one, as systems shift from answering questions to acting: extracting structured cases, routing them, classifying variants, and grounding recommendations in guidelines and case memory.
  3. In 2026, a hematology agent achieved roughly 83% concordance with tumor-board decisions in a prospective silent trial, with hallucinations in 0.3%, suggesting that for well-structured tasks the binding constraint is shifting from accuracy to governability.

The gain

In a prospective silent trial, a hematology AI agent matched tumor-board decisions about 83% of the time, with hallucinations in 0.3% of cases.

The full story

Generative artificial intelligence in hematology is entering a new phase. The dominant question, whether large language models are accurate enough for clinical decision support, is being overtaken by a harder one, as systems shift from answering questions to acting: extracting structured cases, routing them, classifying variants, and grounding recommendations in guidelines and case memory.

In 2026, a hematology agent achieved roughly 83% concordance with tumor-board decisions in a prospective silent trial, with hallucinations in 0.3%, suggesting that for well-structured tasks the binding constraint is shifting from accuracy to governability. Moving from experimental support to reliable infrastructure requires four changes: validating trajectories rather than answers; allocating autonomy inversely to case complexity; treating deployed models as regulated instruments under continuous surveillance; and engineering against automation bias, with data-privacy and consent safeguards throughout.

Sources

  1. Peer-reviewedTurkish Journal of Hematology2026-09-21

The debate