AI in drug discovery: what 24 records show

Every measured gain and problem on the record for this topic, newest first. Updated November 11, 2026.

The count24 counted · 20 context

One square per record, darker for stronger evidence:Evidence: 0 high15 moderate9 emerging

The records

AI in drug discovery

What 24 records show

The records

AI in drug discovery: what 24 records show · Page 2 · Truvace

Where the evidence stands

The short version

Most evidence on AI in drug discovery comes from lab benchmarks, cell assays and animal studies, with very little from patients. TRV-2026-0893 TRV-2026-0940 TRV-2026-2408 Authors of a Nature Reviews Drug Discovery Perspective say that after a decade, evidence of safer, more effective medicines reaching patients faster is disappointingly limited. TRV-2026-0733

What is well measured

A randomized, placebo-controlled trial in 51 healthy adults shows the AI-optimized COVID antibody GB-0669 caused no dose-limiting toxicities and had a 54-day half-life. TRV-2026-0225 In a controlled comparison, 85% of 100 AI-optimized antibiotic peptides were hits, and they outperformed their templates in two mouse models of Acinetobacter baumannii infection. TRV-2026-0193 On trial paperwork, the TrialScout tool linked registrations to published results with 92.5% sensitivity, and human error explained 123 of 200 reviewed disagreements with earlier matches. TRV-2026-0918

Where it falls short

An AI literature screener scored an F1 of 0.684 on drug-drug queries, against 0.872 for drug-family queries, in a test on 7,751 articles. TRV-2026-1610 A GPT-4.1 tool reading consent forms matched experts on 92.1% of forms after tuning, which leaves a share of answers wrong. The authors note that hallucinations limit reliability in high-stakes settings. TRV-2026-1539 An AI tool that masks patient identifiers reached 88.48% average accuracy, which means some trial documents were not correctly desensitized. TRV-2026-1530 A review in the International Journal of Pharmaceutics says no AI-designed polymeric-lipid nanoparticle for breast cancer has entered a registered clinical trial. TRV-2026-1298

Where the evidence is thin

The anticancer prediction model was validated externally on only 10 FDA-approved compounds, and its ranked predictions for 1200 drugs are unconfirmed. TRV-2026-1723 TrialTriage matched the answer key on all 90 synthetic cases, but the authors say real-world performance is not shown. TRV-2026-0666 The AI-designed bird flu vaccine protected mice and ferrets in an animal study, and it still has to be tested in human clinical trials. TRV-2026-2408 An AI framework flagged 817 terminated trials that may have failed from unwanted drug spread in the body, though no drug has been shown revived. TRV-2026-2336

The main threads

Candidates tested in animals and people
5 gains · 0 problems
The strongest results are a small human safety trial of an AI-optimized antibody and animal studies of peptides, a vaccine and prostate cancer drugs. TRV-2026-0225 TRV-2026-0193 TRV-2026-2408 TRV-2026-1644
See these 5 records
Computational screening and benchmarks
4 gains · 0 problems
Models rank well on benchmarks and turn up lab hits, but the work is preclinical and some methods are not built for large screens. TRV-2026-0940 TRV-2026-1348 TRV-2026-1663 TRV-2026-1723 TRV-2026-1954
See these 4 records
Toxicity and safety prediction
1 gain · 1 problem
Lab tests plus machine learning flagged liver-toxic drugs in a 170-drug test, while reviews say black-box toxicity models struggle to win regulators' trust. TRV-2026-1295

The count

Gains: 14
Traces: 2
Problems: 8

A Trace is one record with both a gain and a problem, so it counts on both sides: 16 records show a gain and 10 show a problem. 20 context records are in the record list but not counted, because they report no measured gain or problem.

How strong the evidence is

Evidence	Gain side	Problem side
HighStrong study design, backed by at least two independent sources	0	0
ModerateA measured study or an official source	12	3
EmergingEarly, small, unreviewed or reported second-hand	4	7

By type of evidence: 9 benchmark or lab tests; 6 controlled comparisons; 6 commentaries or narrative reviews; 1 news report; 1 investigation by reporters; 1 randomized trial. How evidence is scored

Strongest gains

AI matched 6,110 of 9,600 trials to published results; humans erred in most disagreements checked
Against earlier human-coded matches, the TrialScout AI tool linked trial registrations to published results with 92.5% sensitivity and 81.2% specificity, and human error explained 123 of 200 reviewed disagreements.
Controlled or validated comparison · Evidence: Moderate · 70/100 · TRV-2026-0918
In a 9,000-sample check, AI masked patient IDs at 88.48%; 99.40% on serious adverse event files
In a check of 9,000 randomly chosen samples from 100,000 processed trial image records, an AI tool masked patient identifiers with 88.48% average accuracy, reaching 99.40% on serious adverse event documents.
Controlled or validated comparison · Evidence: Moderate · 70/100 · TRV-2026-1530
AI-optimized COVID antibody was well tolerated in 51 healthy adults, a randomized trial found
In a placebo-controlled trial of 51 healthy adults, the AI-optimized antibody GB-0669 caused no dose-limiting toxicities, had a 54-day half-life, and showed dose-dependent virus neutralization in blood samples.
Randomized controlled trial · Evidence: Moderate · 70/100 · TRV-2026-0225

Strongest problems

AI drug interaction screener scored lowest on drug pairs in a 7,751-article test
In a test on 7,751 articles, the AI literature screener's F1-score was 0.684 for drug-drug queries and 0.731 for single-drug queries, against 0.872 for drug-family queries.
Benchmark or lab evaluation · Evidence: Moderate · 55/100 · TRV-2026-1610
Big tech says AI can find a cure for cancer. So where is it?
With her liver failing, her skin was a ghostly apricot. The thoughts creep in as we are bombarded by artificial intelligence chief executives making bold proclamations such as “AI will cure cancer in five to 10 years” or “AI will cure cancer in our lifetimes”.