The Udemy Responsible AI Principles were last updated on February 7th, 2025.
This page includes the Udemy Responsible AI Principles.
You can find translated versions of the Udemy Responsible AI Principles by following our Help Center language instructions. The following translations are also available for languages not supported by the Help Center:
Any version of the Udemy Responsible AI Principles in a language other than English is provided for convenience and you understand and agree that the English language will control if there is any conflict.
Udemy Responsible AI Principles
At Udemy, we recognize the transformative potential of Artificial Intelligence (AI) and the opportunities it offers to our learners, instructors, and organizations. However, we also acknowledge the important challenges it presents, which must be carefully addressed in a thoughtful, effective, and compliant way. These AI principles describe our commitment to responsibly developing and deploying AI technologies. We are equally mindful of the challenges AI can pose, which require careful, effective, and compliant solutions. These AI principles reflect our dedication to responsibly advancing and utilizing AI technologies at Udemy.
Data Privacy and Security
We protect user data by maintaining formal security controls and restricting data usage within our AI applications.
Udemy maintains robust controls for secure development practices, data handling, access management, and risk management to protect the confidentiality, integrity, and availability of our systems, services, and data. We continuously monitor our systems and address issues in a timely manner based on severity and risk.
Our policies help protect personal data from unauthorized disclosure:
- For Udemy Business organizations, we do not use personal data, input, or output to train or fine-tune GenAI models for Udemy's GenAI features.
- For individual learners on the Udemy marketplace, we follow the data usage guidelines described in our Terms of Use.
- For instructors, we do not train or fine-tune GenAI models on submitted content for instructors who have opted out of our GenAI program (see Instructor GenAI Policy).
Ethical Alignment
We utilize AI only in ways that uphold our commitment to ethics and social responsibility.
When developing AI applications, we prioritize the rights and interests of our learners, instructors, and organizations. We foster product innovation while maintaining trust by carefully evaluating the possible benefits and risks of AI capabilities for all stakeholders. Our processes include reviewing use cases and mitigating identified biases when designing our systems, which are intended to promote fairness. We implement safety guardrails aimed at preventing inappropriate output. We do not use AI when we cannot adequately resolve ethical concerns.
Transparency
We prioritize transparency when incorporating AI into our products and document design considerations.
Before introducing new product features that incorporate AI, we assess potential impacts and document our technical approaches. We make it evident to users when our products use AI and document terms and conditions of usage where appropriate.
Our AI services are designed to support learning and development. They are not intended nor should they be used for high-risk applications identified by applicable laws, such as employment decisions. Those interacting with our AI capabilities are advised to review and verify any output before relying on it. Udemy Business organizations may access and use the AI features and output, subject to their service agreements with Udemy, solely for supplemental learning and training for their users. Licensing and ownership details for individual learners on the Udemy marketplace and instructors are described in our Terms of Use and Instructor GenAI Policy.
Vendor Compliance and Governance
Our data and AI governance framework requires that internal teams and AI vendors adhere to security, data protection, and ethical AI standards.
We perform security, legal, and IT reviews to assess whether vendors meet our security, data protection, and ethical standards and adhere to applicable data privacy, AI, and other regulations. Vendors must demonstrate necessary controls against algorithmic bias, inaccurate outputs, and intellectual property infringement. Our policy prohibits vendors from training AI models using learner, instructor, and Udemy Business organization data. Our governance programs are designed to help make sure that both internal teams and external partners meet our responsible AI requirements on an ongoing basis.