Will biostatisticians be replaced by AI?
Updated
Biostatisticians score 60 of 100 for AI exposure. The largest share of their task time, 76%, is work AI speeds up while they stay in charge. AI can take over 13% outright, and 11% stays with people.
- 13% AI can do it
- 76% AI helps
- 11% Needs a person
High exposure
Higher than 91% of the 923 occupations we scored.
Task by task
We graded 25 tasks, listed here from most to least exposed. AI can do 3 of them today, speeds up 19 and leaves 3 to people.
| Task | AI exposure |
|---|---|
| Prepare tables and graphs to present clinical data or results. Code generates tables and plots quickly; human checks accuracy and formatting. | 80% AI can do it |
| Write program code to analyze data with statistical analysis software. Coding assistants write and debug R or SAS code well. | 78% AI can do it |
| Write detailed analysis plans and descriptions of analyses and findings for research protocols or reports. Drafting analysis plan text is largely routine writing AI handles well. | 72% AI can do it |
| Prepare articles for publication or presentation at professional conferences. AI drafts much manuscript text; authors take responsibility for claims and revisions. | 65% AI helps |
| Analyze clinical or survey data, using statistical approaches such as longitudinal analysis, mixed-effect modeling, logistic regression analyses, and model-building techniques. Code agents run standard models; statistician chooses and validates specifications. | 62% AI helps |
| Develop or implement data analysis algorithms. Algorithm design is coding-heavy and iterative; human directs and tests. | 62% AI helps |
| Analyze archival data, such as birth, death, and disease records. Archival extraction and analysis automate well; linkage needs human checks. | 62% AI helps |
| Write research proposals or grant applications for submission to external bodies. Grant drafting is heavy writing; funding strategy stays with researchers. | 60% AI helps |
| Draw conclusions or make predictions, based on data summaries or statistical analyses. LLMs can draft interpretations; statistician validates conclusions and uncertainty. | 55% AI helps |
| Calculate sample size requirements for clinical studies. Formulas and software handle calculations; assumptions need expert judgement. | 55% AI helps |
Show the other 15 tasks
| Task | AI exposure |
|---|---|
| Prepare statistical data for inclusion in reports to data monitoring committees, federal regulatory agencies, managers, or clients. Regulatory summaries are structured, but submissions need accountable sign-off. | 50% AI helps |
| Design or maintain databases of biological data. Schema design and maintenance are code-heavy but need domain oversight. | 50% AI helps |
| Design surveys to assess health issues. AI drafts questionnaires quickly; expert validation and piloting required. | 50% AI helps |
| Review clinical or other medical research protocols and recommend appropriate statistical analyses. AI can review protocols and suggest methods; expert must take responsibility. | 45% AI helps |
| Develop or use mathematical models to track changes in biological phenomena, such as the spread of infectious diseases. Model code is automatable; choosing and interpreting biological models stays expert. | 45% AI helps |
| Monitor clinical trials or experiments to ensure adherence to established procedures or to verify the quality of data collected. Data quality checks automate well; on-site monitoring needs people. | 40% AI helps |
| Read current literature, attend meetings or conferences, and talk with colleagues to keep abreast of methodological or conceptual developments in fields such as biostatistics, pharmacology, life sciences, and social sciences. AI summarizes papers; attending meetings and networking stay human. | 35% AI helps |
| Teach graduate or continuing education courses or seminars in biostatistics. AI can draft course materials; live teaching and assessment remain human. | 35% AI helps |
| Provide biostatistical consultation to clients or colleagues. Consultation is live and relationship based; AI helps with preparation only. | 30% AI helps |
| Determine project plans, timelines, or technical objectives for statistical aspects of biological research studies. Timelines and objectives depend on stakeholder negotiation and resources. | 30% AI helps |
| Apply research or simulation results to extend biological theory or recommend new research projects. Extending theory needs expert insight; AI can brainstorm options. | 25% AI helps |
| Design research studies in collaboration with physicians, life scientists, or other professionals. Collaborative design is live discussion; AI only helps with drafting. | 22% AI helps |
| Plan or direct research studies related to life sciences. Directing studies depends on leadership and stakeholder judgement. | 18% Needs a person |
| Collect data through surveys or experimentation. Surveys and experiments need physical or live data collection. | 18% Needs a person |
| Assign work to biostatistical assistants or programmers. Delegating and supervising staff is interpersonal management work. | 12% Needs a person |
How biostatisticians compare
All occupations (923)
Computer and mathematical (36)
Pay and outlook in the US
- Median pay
- $105,650 a year, May 2025
- Projected change in jobs
- +11.0% from 2024 to 2034
Skills to build
- Validate AI-written SAS and R code against source data
- Fit mixed-effects and logistic regression models in SAS or R
- Write statistical analysis plans for clinical study protocols
- Give consultation to clinicians on model choice and interpretation
- Supervise programmers who produce tables and graphs in SAS
Safer roles nearby
5 related occupations with lower scores.
Questions about biostatisticians and AI
Will AI replace biostatisticians?
Partly. About 5 in 10 working hours go to tasks AI can already do, mostly building tables and graphs and writing analysis code in R or SAS. The work that stays with people is choosing the model, checking that the output is correct, and deciding what the results mean for the study.
Which tasks AI is taking over in biostatistics?
Preparing tables and graphs, writing analysis code, and drafting analysis plans have the highest exposure. Code generators produce standard tables and plots quickly, and coding assistants write and debug R and SAS programs well. The statistician still checks each output for accuracy and formatting before it goes into a report.
Which tasks AI is taking over in biostatistics?
Preparing tables and graphs has the highest exposure score at 80, so AI already does much of this work. Code generators produce standard tables and plots quickly. Writing analysis code in SAS or R scores 78, since coding assistants write and debug these programs well. The statistician still checks each output for accuracy and formatting before it goes into a report.
How can biostatisticians stay valuable as AI writes code?
Take responsibility for the analysis specification and check every output that AI produces. Validate AI-written SAS or R programs through independent double programming, and document each model choice in the statistical analysis plan. Spend the time saved on consultation with clinicians and on designing studies before data collection.
Related jobs
Next steps for biostatisticians
Check which roles in your team AI can take on
We map your team's tasks, show what AI can do today and build the automations that pay back first.
Related topics
- Jobs AI can't replaceCement masons score lowest, at 8.
- Jobs most at risk from AIProofreaders score highest, at 76.
- High-paying jobs with low AI exposureRadiologists earn a median $420,860 and score 46.
- Entry-level jobs with low AI exposureCement masons need little prior training and score 8.
- Jobs AI speeds up the mostFor natural sciences managers, AI speeds up 96% of task time.
- AI and jobs in Malta15 iGaming roles graded task by task.
- Skills that keep work humanThe skills to build in the jobs AI changes least.
- Jobs A to ZEvery job we publish, by first letter.
- AI exposure by industryAverage scores for 22 occupation groups.
- Write a CV screening software can readHow applicant tracking systems read a CV, and what to change.
- AI for businessesThe tasks to automate first in each team and industry.
- How we score AI exposureThe grading, the weights and every source.