Using AI responsibly

We integrate AI to empower creativity, protect our community, and uphold high standards of safety and reliability.

Instagram Reel Maker Online No Watermark

How We Use AI at CapCut

End-to-End Intelligent Solutions

Our AI suite makes content creation easy for everyone with rapid video editing tools, AI-generated visuals, virtual avatars, and adaptive music. We enable creators to instantly adopt trending styles and produce high quality results seamlessly with little editing experience needed.

Driving Smart, Effortless Video Production

We leverage advanced AI to turn ideas into ready-to-use videos. Our AI tools simplify production workflow by understanding users' creative ideas, suggesting edits and inspiring video curation—helping creators focus more on storytelling and less on technical complexity.

Keeping Our Community Safe

We prioritize a responsible approach to AI across all stages of development. Our models and features undergo testing before launch and continuous monitoring afterward. By combining human expertise with automated safeguards, we work to maintain a creative environment that’s safe, transparent, and trustworthy for everyone.

Putting Our Commitments into Practice

At CapCut, we prioritize the responsible use of AI by focusing on transparency, fairness, privacy and user safety. Our teams embed these values into every AI-driven feature at CapCut—from content moderation to creative editing tools. Dedicated teams of experts guide how we apply these values in practice through safety and privacy strategies across the platform. Below, we share how these commitments are put into action.

Generate custom AI portraits in seconds, with one click

Transparent AI by Design

Our commitment to trustworthy AI is reflected in our practice of proactive transparency. We inform users when they use AI-powered features and work to provide labels or watermarks to ensure users are informed when content is generated or assisted by AI. Our models and features undergo comprehensive assessments to identify and mitigate emerging risks, including the detection of AI-generated content. These efforts are reinforced by our content moderation framework, which combines human expertise with automated systems to maintain transparency and uphold community integrity.

Preserve original details but generate new style

Prioritizing Safety in Every AI Experience

We are committed to fostering a safe community, respecting diversity, and working to ensure all creators and communities have a positive and fair experience. We continuously evaluate our models and features to detect and reduce risks related to harmful content and bias. Our AI systems adhere to safety guidelines and fairness standards, supported by product safety features, testing, and continuous monitoring to promote consistent outcomes across user groups. Safety and fairness considerations are embedded throughout the product development lifecycle to ensure respect for every user.

Building AI with Privacy in Mind

We embed privacy-by-design throughout the development and deployment of our AI technologies. This includes assessing AI tools for compliance with data protection standards and being transparent about how we process data. CapCut's Privacy Center provides information on our privacy practices and related security details.

Frequently Asked Questions

Does CapCut use generative Al?

Yes, CapCut uses generative AI in a range of creative tools, including AI-generated visuals, virtual avatars, AI-assisted video editing, and other features. These features are designed to make content creation easier for everyone, and they're developed in line with CapCut's commitments to transparency, safety, fairness, privacy, and building trustworthy AI.

How does generative AI work?

Generative AI is powered by sophisticated models trained on millions to billions of data points from sources like text, images, video, and audio. These models use their training data to learn patterns and understand how language, images, video, and audio typically look and sound, which enables them to generate new content in response to questions or instructions.

Different types of generative AI models require different kinds of training data. For example:
- Text models are generally trained on large collections of text and code to learn grammar, structure, facts, and writing styles.
- Image models are generally trained on images paired with text to learn about objects, scenes, styles, and visual concepts.
- Video models are generally trained on video clips and motion information to learn how things move over time, as well as visual style and composition.
- Audio models are generally trained on audio recordings and related information to learn about speech patterns and sound characteristics.

When CapCut offers AI-powered features, those capabilities are supported by models that were trained using a broad mix of publicly available information, proprietary datasets, and synthetic data, consistent with industry practice. In some cases, these datasets may include personal or aggregated information about individuals. CapCut integrates and deploys these models within its products after evaluating their performance and safety so that our features function effectively for users.

How does CapCut prioritize safety and privacy?

CapCut is committed to prioritizing safety and privacy when we develop AI-powered features. We process and protect personal information in compliance with our Privacy Policy and applicable laws. CapCut's Privacy Center provides more information on our privacy practices and related security details.