Responsible AI

Our goal at ByteDance is to build best-in-class tools and technologies that inspire creativity, enrich life, and empower creators, businesses, and everyday users around the world to thrive.

Our AI Tools

ByteDance's AI-powered products and solutions help unlock creativity, enhance productivity, and empower business innovation.

Our AI products come with baseline standards that include:

• Safeguards like minimum age requirements for consumer products, privacy and intellectual property protections, and rules against harmful content like misinformation, nonconsensual sexual content, hate, violence, and minor safety risks.
• Transparency for AI-generated visual content including C2PA Content Credentials, invisible watermarking, and visible AI labels.
• Data storage in secure, compliant infrastructure across supported operating regions.

Each product builds on those standards by adding layers of safeguards for its specific purpose and community. These can include:

• Product-specific privacy protections, safety policies and content moderation technologies.
• Built-in guardrails that block people from misusing each product's features.
• Ways for people to report harmful content, nonconsensual use of likeness, intellectual property violations, or other harms.
• Customizable content filters for certain audiences, like enterprise clients.

Our AI Models

ByteDance builds proprietary text, image, audio, video and coding AI models to power our AI products. Trust & Safety professionals around the world work on model and product safety. This includes:

• Policy and content moderation teams who apply dedicated safety policies to each product.
• AI and safety experts who test, adapt, and build safeguards into both models and features.
• Cross-ByteDance teams who monitor product launches and scale learnings and safeguards across ByteDance's other global AI products.
• Engineers who build transparency technologies like invisible watermarks.

Our Partners

We collaborate with third-party experts throughout model development and deployment, including policy development, testing models for safety or security issues and industry benchmarks.

* listed in alphabetical order