AI News
All stories
Reported

Salesforce adds evaluations, spending controls and a kill switch to Agent Fabric

Salesforce's Agent Fabric adds evaluations, security scans, spending controls and triggered interventions for autonomous agents.

Products · · 1 min read

Salesforce is adding evaluations, security scans, spending controls and triggered interventions to Agent Fabric, its platform for autonomous agents. One of those pieces is a kill switch that can stop an agent when a trigger fires. Agent and API evaluations enter beta in November, while the triggered kill switch is scheduled for general availability in January 2027, roughly two months later.

The announcement does not say what conditions trigger the kill switch, whether it stops one agent or a whole fleet, or who can invoke it. It also leaves out what the security scans check, how the spending controls set limits, and how the evaluations score an agent's behavior. Until those specifics are published, the announcement describes categories of control rather than how each one works.

What changed

Agent Fabric adds agent and API evaluations, security scans, spending controls and a triggered kill switch for autonomous agents.

Type: New capability · Applies to: Agent Fabric

Why it matters

Enterprises running autonomous agents on Salesforce get built-in controls to halt them, but the kill switch will not be generally available until January 2027.

What you can do

Plan agent governance rollouts around the November beta and the January 2027 kill-switch date.

Original source

We have no link to the original source for this story yet.

Where we saw it

1 source

Forward Future (Matthew Berman)

AI wrote this summary from the sources listed, and nobody reviewed it individually before publication. Check the original source before you rely on it. How it works