How do you ensure learner safety in AI conversations?

Learn more about the safeguards GenLearn has put in place to ensure learners can safely interact with AI instructors.

• All GenLearn AI instructors include behavioural guardrails and undergo rigorous testing to ensure learner safety
• Learner sessions are anonymous, with no PII collected and no data is used to train AI models

Learner Safety

1. Guardrails: All learner sessions with the AI instructors have behavioural guardrails and hard cutoffs after a set number of turns, to ensure that learners cannot push the AI to share offensive or harmful information. 
2. Extensive Testing: All AI lesson structures are extensively tested (or 'red-teamed') to ensure the AI does not share offensive or harmful information, and respond assertively to deter these requests from humans.
3. Monitoring: GenLearn conducts human in the loop monitoring for quality control purposes to ensure the AI adheres to its behavioural guardrails and interacts safely with learners.

Learner Privacy

1. Anonymity: All learner sessions are anonymous, and no inputs are used to train AI models
2. No PII: GenLearn does not solicit personal information that could be used to identify learners
3. Data Deletion: GenLearn retains anonymous logs for only 90 days for quality control purposes