Autodesk AI Transparency Cards Come to Its Assistant

Autodesk has brought its AI Transparency Cards into Autodesk Assistant, so users can ask how a specific AI feature uses data and get an answer linked to the relevant card. The company announced the update on October 5, adding a conversational route to information that previously required users to find through the Autodesk Trust Center.Autodesk’s announcement

That makes transparency information easier to reach in the moment a question arises. But the announcement describes an update to documentation and access, not a change to how Autodesk’s AI features process customer data.

What the expanded cards cover

Autodesk launched the cards in 2024. The expanded versions add separate sections on data handling, customer choice and control, model providers, and safety measures. Autodesk Assistant is available across multiple Autodesk products, and its answers identify the source card so users can inspect the underlying information without separately navigating to the Trust Center.The update and its linked-card workflow

The link matters: an answer in a chat interface can summarize an issue, while the source card gives users a place to check the feature-specific information behind it. That is a useful distinction for product teams reviewing AI use in a design workflow. A general statement about a company’s AI policies may not answer what happens with a particular feature; the cards are intended to organize information around those features.

Autodesk says research it conducted with Deloitte partly informed the update. The research indicated that users wanted more detail on training methods and safety measures.Autodesk’s account of the research The announcement does not provide the research methods or results, so it supports the narrower point that Autodesk says those topics emerged as user concerns, not a broader claim about how common or representative those concerns are.

Better access is not a new data control

The new sections include customer choice and control, but their presence in a card does not by itself establish what options users can exercise or whether any controls have changed. Likewise, adding information about model providers or safety measures does not show that the providers, safeguards, or training practices are new.

For teams assessing a particular Autodesk AI feature, the practical improvement is a shorter path from a question to the relevant documentation. The value of that path will depend on how specifically each card explains data handling and user choices. The update makes those answers easier to find; it does not, on the available evidence, change the answers themselves.