Trustworthy, Explainable, and Accountable: How to Give AI Autonomy Without Letting It Run Wild
Salesforce's recipe: probabilistic models + deterministic logic = agents that act without breaking guardrails.
Why it matters
Kathy Baxter lays out a technical pattern for agent autonomy with explainability and accountability built in — pairing learned behavior (probabilistic) with hard rules (deterministic) to keep agents bounded. Relevant to practitioners building agentic systems who need guardrails in production.
The key facts
10 to knowSalesforce Principal Architect of Ethical AI Practice Kathy Baxter as author
Core pattern: probabilistic models paired with deterministic logic
Focus: agent autonomy + explainability + accountability
No specific product launch, pricing, or deployment numbers disclosed
Published Oct 6, 2026 (aligned with Salesforce Winter 27 release window)
Salesforce Principal Architect Kathy Baxter byline
Pattern: probabilistic model paired with deterministic logic
Focus: agent autonomy + guardrails + explainability + accountability
No pricing, no availability date, no deployment metrics disclosed
Source: Salesforce blog/news story (vendor-authored)
Go to the source
Salesforce Newsroomsalesforce.com
Publisher excerpt: Salesforce Principal Architect of Ethical AI Practice Kathy Baxter explains how pairing a probabilistic model with deterministic logic gives agents the ability to take action — without giving up the guardrails.