A company wants an Agentforce agent to update Case records and execute Flows on behalf of support agents. Which security model ensures the agent does NOT exceed the user's data access?

The correct answer is C. Running the agent in the user context. Running the agent in user context means all Salesforce record access, DML operations, and Flow executions are governed by the permissions of the user on whose behalf the agent is acting. This ensures the agent cannot access or modify records the user themselves could not. System

AI Features for Service (e.g., Einstein Bots, Next Best Action)

Question

A company wants an Agentforce agent to update Case records and execute Flows on behalf of support agents. Which security model ensures the agent does NOT exceed the user’s data access?

Options

ASystem Mode execution
BEinstein Trust Layer only
CRunning the agent in the user context
DPublic API access

How the community answered

(30 responses)

A
13% (4)
B
7% (2)
C
77% (23)
D
3% (1)

Explanation

Running the agent in user context means all Salesforce record access, DML operations, and Flow executions are governed by the permissions of the user on whose behalf the agent is acting. This ensures the agent cannot access or modify records the user themselves could not. System Mode execution (A) runs with elevated, admin-level permissions that bypass user-level restrictions and could expose or modify data beyond the user's entitlement. Einstein Trust Layer (B) governs AI data security and masking, but does not enforce Salesforce record-level access control. Public API access (D) is not a valid Agentforce security model.

Topics

#AI Agent Security#User Context#Salesforce Security#Agentforce

Community Discussion

No community discussion yet for this question.

Full AI-201 Practice