Recommended Free Tools
iTechGuides is reader-supported. When you buy through links on our site, we may earn an affiliate commission. As an Amazon Associate I earn from qualifying purchases. Learn more
AI can perpetuate racial bias in cancer care when its training data or measurements do not represent the people it is used to help. But the evidence cited here does not establish that a deployed cancer AI system denied a minority patient life-saving treatment. The distinction matters: unequal representation creates a real risk to investigate, not proof that a particular algorithm caused a treatment denial.
Can AI be biased in cancer treatment?
Yes. A model learns patterns from the data it is given. If some groups are underrepresented, have more missing information, or receive tests measured differently, predictions may be less reliable for them. A mismatch between the people represented during development and those seen in practice can create another source of error.
The National Cancer Institute (NCI) put the risk plainly on its May 30, 2024, page “Artificial Intelligence (AI) and Cancer”: “However, if data used to train AI models are not appropriately diverse and representative of the broader population, these models can perpetuate medical bias.” NCI also notes that AI could help extend access to cancer care. Its potential benefits and risks depend on the particular tool and how it is used.
A prediction may inform cancer risk assessment, diagnosis, referral to a clinical trial, or a treatment decision. If people rely on a flawed prediction, it could contribute to unequal care. That pathway is plausible, but an imbalanced dataset alone does not show that a model caused an adverse decision or changed a patient’s outcome.
What do the available examples show?
Representation is a documented concern, but the figures below describe particular datasets and studies—not every cancer AI system. The studies also address different tasks, so their results should not be treated as a direct performance comparison.
| Evidence | What it reports | What it can establish |
|---|---|---|
| TCGA figures in a presentation by Faisal Mahmood, Ph.D., hosted by the U.S. Food and Drug Administration in 2024 | European-ancestry patients accounted for 82.0% of all cases in TCGA. In the presentation’s TCGA lung adenocarcinoma cohort, 566 people were listed; 68.6% were White, 9.2% Black, and 1.4% Asian. | These specific figures illustrate representation concerns in TCGA. They are not a measure of the demographic makeup of all cancer AI datasets, nor do they alone show that a model produced biased care. |
| 2025 multimodal-model study using five randomized phase III prostate cancer trials | The evaluation included 5,708 patients. The study reported performance across African American and non-African American subgroups without evidence of algorithmic bias. | This is bounded evidence about that model, those data and endpoints, and those subgroups. It does not establish that other models are unbiased or that performance is equal in every setting. |
The examples answer different questions: one describes who was represented in specified research data, while the other evaluates a particular model across subgroups. Neither by itself establishes whether a system improved or restricted real-world access to treatment.
Rank #2
- Track important events like birthdays, anniversaries, holidays, and other special occasions so you don’t forget.
- Keep up with activities such as meetings, dates, appointments, and deadlines.
- Organize work projects and tasks to stay on track throughout the year.
- Jot down ideas and keep a running “to do” list of things you want to accomplish.
- Set reminders for yourself to motivate yourself to reach your goals.
Does cancer AI work as well for Black patients?
There is no single answer for cancer AI as a whole. It depends on the model’s purpose, its development and evaluation data, the quality and completeness of those data, and the clinical setting in which it is used. The 2025 prostate cancer study offers evidence about one model and defined subgroups; it cannot guarantee performance for Black patients using a different tool, for another cancer, or in another care setting.
“Fairness” also has several distinct meanings. Researchers and clinicians need to ask whether a model predicts accurately for each group, whether its risk estimates are calibrated, whether patients have comparable access to the care connected to those predictions, and whether using the model improves patient outcomes. Similar measured performance on one evaluation does not answer all of those questions.
Why removing race is not a universal fix
Leaving race out of a model does not automatically make it fair. A 2024 Proceedings of the National Academy of Sciences study on colorectal cancer risk prediction found that race adjustment could improve performance when family-history data quality varied by race. The Agency for Healthcare Research and Quality’s evidence review, “Impact of Healthcare Algorithms on Racial and Ethnic Disparities in Health and Healthcare,” also addresses the effects of algorithms on disparities.
The practical question is whether the model uses appropriate features and measurements for its specific task, and whether its predictions have been evaluated against suitable outcomes. A variable may encode differences in data collection or access as well as clinically relevant information. Deciding whether to include or exclude it requires examining how it affects performance and decisions—not applying a blanket rule.
Rank #4
How do researchers test AI for racial bias?
A useful evaluation looks beyond a single overall accuracy figure. Researchers should examine who is represented, how complete and reliable the measurements are, and whether the model’s performance holds for relevant groups in the setting where it will be used.
The Tool Desk
Outbyte Driver Updater FREEFix the driver behind crashes, sound loss and screen glitchesFind Drivers →Outbyte PC Repair FREEClear out junk files and repair common Windows errorsFree Scan →- Define the task and outcome. A model that estimates risk is not performing the same job as one intended to predict drug response or support diagnosis. The intended use determines what counts as a meaningful error.
- Check development and evaluation cohorts. Report the populations represented, subgroup sizes, missing data, and measurement quality. A headline sample size can conceal that a subgroup is too small for a reliable estimate.
- Measure performance by group. Assess predictive accuracy and calibration separately where the data support it. These measures answer different questions and do not, on their own, establish fair access or better outcomes.
- Test in the intended care setting. A retrospective result may not carry over to a different hospital, patient population, workflow, or decision. NCI identifies further randomized trials as one need for validating clinical uses.
- Monitor clinical use and patient effects. Check whether model-supported referrals or decisions work as intended in practice, including whether they affect access to care. A model’s behavior can be shaped by the workflow around it, not only by its code.
NCI also identifies broadly accepted standards and explainable AI as important needs for integrating tools into clinical workflows. No single metric substitutes for checking the data, decision process, and outcomes together.
Best Value
- [Pink Ribbon and Inspirational Design] This medical logbook centrally organizes all your breast cancer health records, including follow-up records, medication plans, post-operative recovery, and breast-related medical history. Say goodbye to disorganized documents
- [Premium Materials]Features a thick, durable hard cover with robust double-thread binding, built-in category labels, and convenient elastic bands. Every detail has been meticulously designed for both sturdiness and durability, while also ensuring easy everyday browsing and carrying
- [Convenient Design] 6.5 x 8.4 inches, thick hardcover, double-stitched binding, divider labels, elastic band, pen holder. Thick, ink-resistant inner pages, three-section bookmarks, and inspirational coloring pages provide daily encouragement
- [Comprehensive Layout] Divided pages designed specifically for breast cancer patients, with dedicated sections for breast cancer medical records, surgical records, chemotherapy and radiotherapy records, follow-up records, and emergency care records. This allows you to quickly find key medical information during follow-up visits, emergencies, or other unexpected situations
- [Meaningful Health Gift] The pink ribbon theme conveys support and care, making it a meaningful gift for yourself or for friends and family experiencing or overcoming breast cancer. Add a touch of warmth to breast cancer awareness month
How AI research relates to cancer-treatment access
Algorithmic bias is one concern; barriers to research and care are another. NCI describes Black and Latino patients as underrepresented in NCI-supported clinical trials and lists the ECOG-ACRIN SUPPORT trial, which tests navigation and research-literacy approaches to diversity participation and enrollment. This work addresses trial participation; it does not show that AI caused underrepresentation or denied anyone treatment.
That distinction is important when interpreting claims about cancer AI. Access barriers can affect who enters a study, whose data become available, who receives follow-up care, and who can act on a recommendation. Those problems can intersect with model design, but they are not interchangeable evidence.
What cancer AI is still experimental?
NCI’s PERCEPTION project is a proof-of-concept research tool, not an established bedside treatment-selection system. It uses single-cell RNA sequencing to predict drug response from tumor cell populations. Researchers built models for 44 FDA-approved cancer drugs using cell-line data, then evaluated published patient datasets that included multiple myeloma, breast cancer, and non-small cell lung cancer. NCI notes that single-cell sequencing costs more than bulk sequencing and is not yet widely available clinically.
The FDA Oncology Center of Excellence says its Oncology AI Program began in 2023 in response to growing interest in oncology AI. The existence of the program reflects regulatory and research attention; it does not by itself establish that a particular cancer AI system is approved or that a specific regulatory rule applies.
What patients and clinicians should take from this
AI may support cancer research and care, but a system’s presence is not evidence that it is reliable for every patient or that it has improved treatment access. For any tool influencing care, the relevant questions are what decision it informs, which patients were represented in its evaluation, how it performed for relevant groups, and whether its use has been shown to improve outcomes in practice.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
