Artificial intelligence in pathology is shifting from simply detecting cancer to predicting which patients will respond best to specific treatments. The FDA has cleared the first AI tool designed to inform treatment decisions directly from tissue samples, marking a turning point in how oncologists use digital slides to guide therapy .

How Is AI Changing Cancer Treatment Decisions?

In August 2025, the FDA granted approval to ArteraAI Prostate, the first AI-powered digital pathology tool cleared to provide both prognostic and predictive information in localized prostate cancer . Instead of simply answering "Is this patient high risk?", the tool enables oncologists to ask a more actionable question: "Is this tumor biologically likely to benefit from treatment intensification?" . The AI analyzes whole-slide images of biopsies plus clinical variables to estimate a patient's 10-year metastasis risk, prostate cancer-specific mortality, and likelihood of benefiting from treatments like androgen deprivation therapy or androgen receptor pathway agents combined with radiation.

The tool has already gained traction in clinical practice. It is included in the National Comprehensive Cancer Network (NCCN) Prostate guidelines at category 2A, and Medicare payment was established effective January 1, 2024 . The FDA's approval includes a predetermined change control plan, allowing the company to add scanner compatibility without requiring full resubmission, streamlining future updates.

To validate real-world impact, the UK Vanguard Path program will evaluate how well ArteraAI's predictions align with actual outcomes among more than 4,000 men treated for prostate cancer over at least five years, measuring whether the AI's recommendations actually change treatment decisions and timelines .

What Other AI Tools Are Helping Pathologists Catch Cancer?

Before predictive tools like ArteraAI, assistive AI systems are already embedded in pathology workflows as safety nets. Paige Prostate Detect was the first FDA-authorized AI in pathology, flagging suspicious areas on prostate core biopsies and receiving class II designation through the FDA's de novo pathway in 2021 . In April 2025, Paige PanCancer Detect received breakthrough device designation for detecting suspicious foci across multiple tissues and organs, becoming the first multitissue AI assist to earn this designation . The company was acquired by Tempus in August 2025, merging approximately 7 million slides with multiomics data to accelerate the path from discovery to regulated tools.

Ibex Prostate Detect, developed by Galen Second Read, received FDA 510(k) clearance in February 2025 . This tool flags cases initially signed out as benign for re-review, producing case-level alerts and heatmaps showing likely cancer regions. These assistive tools function as pragmatic backstops in high-volume pathology services, reducing false negatives and standardizing triage without replacing human pathologist judgment .

How AI Is Improving Tissue Use and Molecular Testing

A broader shift in oncology AI involves tools that improve downstream clinical workflows beyond detection. Virtual staining represents this evolution. ClearStain, developed by Pictor Labs, generates a hematoxylin and eosin (H&E)-equivalent digital image from an unstained tissue section destined for molecular testing . Pathologists can annotate tumor on the same section that undergoes DNA and RNA extraction, ensuring "what you see is what you sequence." Proscia is integrating Pictor's virtual stains into its Concentriq platform, signaling ecosystem adoption.

The implications are significant for tissue-limited cases. This approach preserves tissue in small biopsies, improves tumor purity selection, reduces failure of next-generation sequencing (NGS) due to tissue depletion, and shortens turnaround time . For tissue-limited lung or prostate rebiopsies, this capability is immediately relevant. The conceptual shift is profound: the diagnostic slide becomes an active guide for downstream molecular workflows rather than a static endpoint.

What's Next: Multimodal AI Models and Genome-Guided Drug Discovery

Companies like Noetik are training multimodal transformer models on paired H&E slides, spatial transcriptomics, spatial proteomics, and sequencing data from thousands of tumors, analyzing nearly 40 million cells . These systems simulate spatial single-cell gene expression in context. While not yet FDA-cleared diagnostics, they represent the pipeline for future computational biomarkers. Partnerships, such as those with Agenus, aim to derive predictive biomarkers for immunotherapy directly from morphology plus learned biology. The trajectory is clear: detection, prognosis, treatment prediction, and pathway inference are all derived from the diagnostic slide .

Steps Oncologists Should Take to Prepare for AI-Integrated Pathology

• Validate whole-slide imaging readiness: Confirm with your pathology group whether primary diagnosis has been validated on the current whole-slide imaging platform according to College of American Pathologists guidance, including appropriate case mix and concordance benchmarks.
• Understand AI tool capabilities and limitations: Learn whether your institution uses detection-assist tools like Paige or Ibex, and how they integrate into your pathology workflow as safety nets rather than autonomous diagnostic systems.
• Explore treatment-predictive tools: Investigate whether ArteraAI Prostate or similar predictive tools are available for your patient population and how they can augment traditional prognostic markers like Gleason grade and prostate-specific antigen.
• Plan for virtual staining integration: Work with your pathology and molecular testing teams to understand how virtual staining could preserve tissue and improve sequencing success rates for tissue-limited biopsies.

Why the Shift From Detection to Prediction Matters

The Moravec paradox, a concept in artificial intelligence, states that what is easy for humans is often hard for AI, and what is hard for humans may be easy for AI . In digital pathology, this dynamic is visible: AI excels at large-scale scanning and quantification, while humans excel at contextual reasoning and clinical integration. The future is not replacement but division of labor .

"Instead of asking, 'Is this patient high risk?' you can begin asking, 'Is this tumor biologically likely to benefit from intensification?' This approach does not replace genomics or clinical judgment. It augments them with slide-derived biology," explained Arturo Loaiza-Bonilla, MD, MSEd, Chief Medical Officer at Massive Bio and Systemwide Chief of Hematology and Oncology at St. Luke's University Health Network.

Arturo Loaiza-Bonilla, MD, MSEd, Chief Medical Officer at Massive Bio

This convergence of validated digital slides, stronger AI backbones, and clearer regulatory guidance sets the stage for the next shift in oncology: AI that doesn't just detect cancer but informs treatment decisions . As these tools move from research into clinical practice, they promise to make cancer care more precise and personalized, helping oncologists match treatments to the specific biological characteristics of each patient's tumor.