Agentforce trust and access controls for Life Sciences agents
Key Takeaway
Permission sets alone don't expose a Life Sciences agent. To let a user use an agent such as Pre Visit Planning, enable the agent for that user. Move it to Enabled Agents under Enabled Agent Access in a permission set assigned to them, or add agent access to their profile.
Agentforce agents in Life Sciences Cloud, such as the Pre Visit Planning agent, work with sensitive customer and health data. Salesforce protects them with two layers. The Einstein Trust Layer controls how data moves to and from the large language model (LLM). Standard Salesforce access controls decide which users can use an agent and what the agent can reach. A consultant needs to know both layers. A consultant also needs to know the permission sets, agent access, flow access and admin settings that Life Sciences agents require.
Standard Salesforce access controls
AI agents respect standard Salesforce access controls, including licenses, permissions, field-level security and sharing settings. You secure an agent with the same security model you already use for users, not with a separate one.
Access to a custom agent action comes from the Apex class, flow or prompt template that the action references. For example, an action built on a flow follows the permissions, field-level security and sharing settings configured in that flow.
Licensing controls access too. The default Agentforce agent is available to users with the "Agentforce Default" permission sets. Some agent types and standard actions are built for specific clouds or products, so they need another license or permission.