Artificial intelligence-generated marathon training programs: reliable tools in exercise prescription for athletic...

Area of Science:

• Sports Science
• Artificial Intelligence in Fitness

Background:

• Marathon running is increasingly popular globally.
• Training requires structured plans for pacing, volume, and periodization.
• The role of AI in prescribing evidence-based marathon training is not yet clear.

Purpose of the Study:

• To evaluate the ability of leading AI models to generate evidence-based marathon training plans.
• To compare AI-generated plans against established endurance training principles.

Main Methods:

• A descriptive study analyzed outputs from eight leading AI models.
• AI models were prompted to create 6-month marathon training plans for beginner, intermediate, and advanced runners.
• Outputs were qualitatively assessed against peer-reviewed marathon training literature.

Main Results:

• Most AI models identified key training elements like mileage progression, tapering, and low-intensity focus.
• Significant variations in accuracy and completeness were observed across AI models.
• Some AI plans lacked detail, failed to differentiate athlete levels, or provided inconsistent pacing data, particularly for advanced runners.

Conclusions:

• AI has strong potential for generating accessible, structured training content aligned with current principles.
• Limitations in personalization and the need for professional oversight require further validation.
• Future research should focus on real-world outcomes and integrating personal physiological data into AI training programs.