Beyond ChatGPT: A Review of the Use of AI Tools in Biological Education

Synthesis: Cotton and Cotton (2026) review the use of AI tools in biological education, noting that AI has revolutionized biology research (exemplified by AlphaFold's Nobel Prize) yet its adoption in biology education has been slower, partly due to concerns about generative AI (GAI) tools like ChatGPT. The review examines the potential benefits of AI — enhanced student engagement and subject knowledge, support for coding skills, assistive technologies for students with disabilities, and predictive modeling to identify at-risk students — alongside emerging specialized tools such as iNaturalist and Google Lens for species identification and Machine Learning tools for bioimaging. Evidence suggests tools like iNaturalist can improve learning outcomes, promote engagement, and foster environmental stewardship. Challenges include academic integrity, assessment design, misinformation, and the potential erosion of critical thinking and independent research skills, requiring professional development for educators and clear guidance for students.

The context

AI has transformed biological research — from AlphaFold's protein-structure prediction (Nobel Prize 2024) to 'omics, molecular/cellular engineering, cancer biology, animal behavior, and biodiversity work. These developments increasingly underpin biology education, making consideration of AI in this context urgent. Prior to 2022, AI use among scientists required high-level programming; ChatGPT's arrival made GAI easy to use and freely available, but also triggered concerns about assessment integrity, misinformation, and erosion of Problem Solving skills. Notably, around 16% of students surveyed in Sweden and the UK admitted using GAI in assessments, and GPT-4-generated essays went almost entirely undetected and outscored undergraduates.

Benefits of AI in biological education

• Enhanced engagement and subject knowledge — AI tools support interactive, personalized learning.
• Support for coding skills — AI assists students with the computational demands of modern biology.
• Assistive technologies — AI supports students with disabilities.
• Predictive modeling — identifying at-risk students for early intervention.
• Species identification and fieldwork — computer-vision tools like iNaturalist and Google Lens identify species from images, supporting learning and engagement, especially in fieldwork. Evidence suggests iNaturalist can improve learning outcomes and foster environmental stewardship.
• Bioimaging and specialized ML — emerging machine-learning tools for bioimaging and species identification in biology teaching.

Challenges and risks

• Academic integrity — GAI easily completes many traditional assessments; detection remains imperfect.
• Assessment design — the need to redesign assessments that AI cannot trivially complete.
• Misinformation and hallucination — AI can produce confident but inaccurate content.
• Erosion of critical thinking and independent research skills — over-reliance on GAI may weaken these.
• Environmental cost — the substantial computing resources behind GAI.

What this means for practice

• Instructors. Reach for specialized, evidence-backed tools as well as chatbots — image-based species identification (iNaturalist, Google Lens), Machine Learning bioimaging tools, and predictive modeling to flag at-risk students — rather than treating GAI as the only option.
• Instructors. Stop relying on detection: GAI completes traditional assessments easily and detection remains imperfect, so redesign tasks that AI cannot trivially complete and assess understanding through them.
• Instructors. Teach students to check AI output for misinformation and hallucination explicitly, so that overreliance does not erode critical thinking and independent research skills.
• Instructors. Put assistive AI supports in place for students with disabilities and build coding support into biology courses, both of which the reviewed evidence ties to engagement and subject knowledge.

Limitations

• The review covers research published between 2020 and 2024 and adds no new primary data, so it can map the field but cannot test any intervention.
• The reviewed corpus is dominated by individual case studies with varied levels and quality of evaluation; the authors find a dearth of large-scale, national or international studies.
• Longitudinal evidence is scarce: the review reports limited research on AI's long-term impact on student retention, comprehension, and career outcomes in biology.
• Several included studies lacked a control group (for example, an AI-enhanced e-book study), and the authors identify fieldwork, authentic assessment, and the environmental cost of AI as gaps in the literature they reviewed.

Citation

Cotton, P. A., & Cotton, D. R. E. (2026). Beyond ChatGPT: A review of the use of AI tools in biological education. Journal of Biological Education. Advance online publication. CC BY-NC-ND 4.0.