Zendesk offers multiple features powered by generative artificial intelligence (AI) to enhance the delivery of exceptional customer service at scale. These features leverage large language models (LLMs) developed by third-party vendors that integrate with Zendesk.
Understanding Generative AI Features
Our platform offers a variety of generative AI capabilities designed to enhance your experience and productivity. Here's a quick overview to help you navigate these features:
1. Copilot Features
Copilot includes multiple generative AI functionalities that assist you across different tasks. All Copilot capabilities are centralized within an AI hub, providing you with a streamlined place to explore, manage, and adopt AI-powered tools effortlessly.
Learn more: About agent copilot | Monitoring and optimizing effective AI setup in your account
2. AI Agents Features
AI Agents leverage generative AI technologies to perform actions and support your workflows. These tools are designed to automate and simplify processes for you.
Learn more: Getting started with AI agents - Essential | Getting started with AI agents – Advanced
3. Other Generative AI Features
Apart from Copilot and AI Agents, we offer additional generative AI features that complement your experience, such as Generative Search.
Protecting Your Data
We’re committed to protecting your information while providing powerful AI features to enhance your experience. In this section, you’ll find important details about how we safeguard your data, the agreements that protect your privacy, and the guidelines for using AI-powered services safely and responsibly. Visit our Trust Center to learn more about our data protection practices and commitments.
1. Contractual Protections
All of the protections contained in your Zendesk Customer Agreement apply to your use of any generally available AI features. If you've entered into a Data Processing Agreement or Business Associate Agreement with Zendesk, those protections will also apply.
Learn more: Advanced Compliance for details on how Zendesk AI can be used in compliance with your HIPAA obligations.
2. Third-Party Contract Restrictions
Zendesk is committed to implementing appropriate terms with our LLM vendors to protect your Service Data, including all required Data Processing Agreements and/or Business Associate Agreements. For any LLM vendors considered Zendesk’s sub-processors, these vendors cannot use your Service Data to train their models.
Learn more: Sub-processor Policy
3. Zero Data Retention
Zendesk integrates LLMs into the Services in two ways:
- Using LLMs which are hosted on Zendesk-controlled infrastructure within AWS Bedrock, Microsoft Azure, or Google Cloud Platform. This means that the LLM provider never accesses your Service Data, and is not a sub-processor of Zendesk.
- Using LLMs directly via API. For these use cases (currently, only OpenAI), Zendesk leverages ‘zero data retention’ endpoints, meaning request and response bodies are not persisted to any logging mechanism and exist only in memory in order to serve the request. No data is stored by OpenAI following the delivery of the output.
4. Data Locality
No LLM provider stores Service Data to provide generative AI functionality – all processing occurs with zero data retention. Zendesk also takes additional steps to secure data residency commitments from our service providers. For example, Zendesk is one of the first companies to access OpenAI’s European endpoints, enabling European customers to leverage generative AI with extra confidence.
Generative AI Transparency
Generative AI features are identified to users within Zendesk. Certain features can be enabled or disabled for each instance, while others will only leverage generative AI if proactively used. Features that leverage generative AI will be disclosed in the user interface, so that customers can make informed choices about whether / how to deploy generative AI.
For features where AI generates responses directly to your customers, Zendesk identifies the AI-generated responses so that your customers are always aware when they are interacting with AI.