Note
Access to this page requires authorization. You can try signing in or changing directories.
Access to this page requires authorization. You can try changing directories.
These frequently asked questions (FAQ) describe the AI impact of the agent creation from natural language experience in Copilot Studio.
What is the agent creation from natural language experience?
The agent creation from natural language experience has two entry points in Copilot Studio: the Home page and the Agents page.
You describe what you want your agent to do in your own words, and Copilot Studio automatically generates a name, description, and instructions for your agent.
What are the capabilities of the agent creation from natural language experience?
If you create an agent by entering a natural language query on the Home page or on the Agents page, the AI generates a name, description, and instructions for your agent. Additionally, you see suggestions for channels and triggers, connectors, and knowledge sources such as SharePoint, Dataverse, and public websites.
What is the agent creation from natural language experience's intended use?
Use this experience to create an agent with some AI-generated starting information to help you build and test your agent more quickly.
How was the agent creation from natural language experience evaluated, and what metrics are used to measure performance?
The evaluation process involved assessing the quality of the generated name and description, instructions, suggested knowledge, suggested triggers, and suggested tools on 250 different natural language inputs on a scale of 0-100. These scores are then aggregated into a weighted metric. The evaluation also measures the rate of acceptance and dismissal of the suggestions for production traffic, and explicit feedback through thumbs up and down mechanisms.
What are the limitations of the agent creation from natural language experience, and how can users minimize the impact of limitations when using it?
This experience doesn't generate suggested prompts, tools outside of prebuilt connector actions and triggers, or knowledge sources outside of SharePoint, Dataverse, and public websites. You can manually configure your agent to add any of these capabilities.
What operational factors and settings allow for effective and responsible use of the agent creation from natural language experience?
There are no operational factors or settings you need to consider to use this experience.
What protections does Copilot Studio have for responsible AI?
Copilot Studio uses content moderation policies on all generative AI requests to protect admins, makers, and users against offensive or harmful content. These content moderation policies also protect against malicious attempts at jailbreaking, prompt injection, prompt exfiltration, and copyright infringement. The system checks all content twice: first during user input and again before showing the AI-generated content and suggestions. If the system finds harmful, offensive, or malicious content, it prevents your agent from responding.