JFrog Responsible
AI Principles

JFrog Responsible AI Principles

At JFrog, we aim to create a balanced and ethical approach to AI to enable our customers to utilize AI within our products, which will be a trusted assistant providing time-saving outputs. By adhering to these Responsible AI Principles, we aim to contribute to the responsible and beneficial development of AI technologies in a fair, transparent and ethical manner.

1. Accountability and Responsibility

Our Principles program starts with a commitment from our cross-functional key executives. Our executive AI committee meets regularly to review and approve aspects of the Responsible AI program and advise on novel issues as they arise. JFrog maintains clear roles and responsibilities for stakeholders involved in the AI lifecycle. We have implemented a robust governance framework that includes impact assessments, data governance policies, and human oversight to ensure accountable AI deployment.

2. Transparency and Explainability

JFrog enables our customers to make informed choices about their engagement with AI systems. The user experience is designed to make it clear when users are interacting with an AI system. 

3. Privacy and Security

JFrog has implemented data protection measures to ensure that AI systems are designed with privacy and security considerations. JFrog doesn’t train models on specific customer data but rather on aggregated data that does not identify JFrog customers or their users.

4. Reliability and Safety

JFrog prioritizes the development of reliable and safe AI systems by conducting comprehensive testing and validation. We have established mechanisms to monitor and mitigate potential cyber threats.

5. Continuous Improvement and Learning

JFrog has fostered a culture of continuous improvement by utilizing adaptive training, feedback loops, user education, and regular compliance reviews. We encourage cross-disciplinary collaboration to enhance the AI system’s capabilities and address emerging challenges.

6. Fairness and Non-Discrimination

AI systems will be designed and implemented with human oversight and diverse perspectives to mitigate the risks of discrimination or harmful bias. Regular audits and fairness checks will be conducted to identify and mitigate any potential bias.

7. Sustainability

JFrog’s AI development principles will strive to minimize the environmental impact of our AI technologies and consider their long-term effects on the planet.

8. Governance

JFrog carefully selects our third-party AI vendors to ensure a commitment from the vendor that they will forgo using JFrog and JFrog customers’ content to develop, train, and fine-tune vendor models.

AI at JFrog FAQs
At JFrog, we aim to create a balanced and ethical approach to AI to enable our customers to utilize AI within our products, which will be a trusted assistant providing time-saving outputs. We have developed these FAQs to provide the information you need to make informed decisions on your use of JFrog AI Tools.

JFrog leverages models developed by its subprocessors listed and updated at: https://jfrog.com/trust/privacy/sub-processors/. You can read more about their safety and AI practices at their respective sites. 

JFrog doesn’t train models on specific customer data but rather on aggregated data that does not identify JFrog customers or their users.

Any information used to power JFrog AI Tools  will be shared with our subprocessors for the sole purpose of providing you with features. We require all subprocessors to agree to contractual obligations that prohibit any use of customer data to train their models. Learn more about JFrog AI security and privacy practices at https://jfrog.com/responsible-ai-principles/.

We test models internally before deployment to our customers to promote accuracy and reliability for our users.

At JFrog, we keep pace with the rapid evolution of AI tools by constantly updating and refining the models we consume.

Yes, we carefully select third-party AI vendors who commit to the same privacy standards we maintain. All our AI vendor partnerships include contractual guarantees that customer content will not be used for model development, training, or fine-tuning. Our AI models are listed and updated on the JFrog Sub-Processors page. 

JFrog teams approach the development and deployment of AI with a focus on a balanced and ethical approach to enable our customers to utilize AI as a trusted assistant providing time-saving outputs. See our Responsible AI Principles. JFrog understands the growing concerns around AI and data usage. JFrog does not use Customer Data, Artifacts, or Binaries to train AI models.

Our user experience is designed to clearly indicate when you're interacting with AI systems. We believe in transparency and ensure you can make informed decisions about your engagement with our AI features.

Your input data is processed in real-time to generate responses and is not stored for training purposes. We implement enterprise-grade security measures to protect your data during processing, and all data handling complies with industry privacy standards.

Yes.

We maintain complete separation between customer data and our AI processes. Your proprietary information, code, documents, and other intellectual property remain confidential and are never used to improve the models or shared with other customers.

We apply continuous monitoring on the “evals” produced by the system to detect any drifts or unwanted behavior.

For questions about our AI practices, data handling, or privacy policies, please contact our support team or privacy@jfrog.com.  We're committed to transparency and are happy to provide additional information about our responsible AI approach.