Sales agent platform Outreach says every agent action gets a full audit trail; runtime enforcement and insurance undisclosed

Outreach, a sales-engagement platform whose AI agents surface prospects and guide sellers on next steps, states that customer-configured limits on what its agents can do come with "full audit trails for every action taken," according to the company's website.

The company describes the controls as "fully configurable," letting a customer define what an agent can and cannot do, and separately says agent access is secured through single sign-on and role-based access controls "built to secure the most sensitive revenue data." Outreach says its platform helps sellers "close over 2 million opportunities" a month — a figure the company reports about itself.

That language describes what a customer can configure and, after the fact, audit. It does not say whether the configured limits are checked and enforced automatically as an agent runs, or only reviewable afterward. aidispatch.news's index records Outreach's runtime-governance field as no public information found, based on an August 5, 2026 capture of outreach.ai.

A more recent capture of the same site — 5,214 characters, fetched September 23 and checked for the quotes above — contains no mention of insurance, a named carrier or indemnification. It also contains no mention of a compliance certification, outcome-based pricing or a settlement mechanism for a failed agent task.

Outreach is a candidate entry in aidispatch.news's index of AI agents and vendors. It has not completed human review on its trust-gap scorecard; an automated, unreviewed pass credits it with partial signal on the runtime-governance dimension, drawn from the single sign-on and role-based access language above — not from the runtime-governance field itself, which the index separately records as no public information found.