AI Dispatch — Daily Dispatch, August 10, 2026

# AI Dispatch — Daily Dispatch, August 10, 2026

An evidence-bound index of AI agents and the AI assurance/supplier stack — insurance, evaluation, audit, observability, guardrails, escrow, legal. Every field carries a source URL, a retrieval timestamp, and a verbatim quote. No estimates, no growth-hack framing.

## Sponsor disclosure

No sponsored placement in this issue. `listings` carries no `featured`/`active` row, checked directly against the database at draft time (August 10, 2026, 16:22 UTC).

## What published on August 10, 2026

No new entity cleared to published status in the 24 hours ending at draft time (August 9, 2026, 16:22 UTC through August 10, 2026, 16:22 UTC). No new field extractions were recorded for existing entities in the same window, and the change-detection queue held zero pending rows.

The day's only new material was three wire articles, all published August 10, 2026 at 12:31 UTC, and all describing the same pattern: a vendor selling governance, audit or automation capability into regulated industries while disclosing nothing about its own insurance or liability terms.

- **DeepOpinion** — an enterprise automation platform (branded "Otera" on its current homepage) that says it processes more than 600,000 regulated cases a year, including "core operations in insurance, finance, and governance that involve complex validation checks, approval workflows, and audit trail maintenance." The company states it holds compliance "with SOC 2 Type II, ISO 27001, GDPR, DORA, HIPAA, and EU AI Act certifications," none independently verified by a published certificate or auditor name. AI Dispatch found no public information on insurance availability, named carriers, coverage limits, indemnification or a dispute process for DeepOpinion's own service — five of the categories tracked under the trust-gap rubric. (full piece: `deepopinion-600000-cases-no-insurance-terms`) - **Fiddler AI** — a governance and observability vendor whose guardrails enforce policy on live agent traffic. The company states: "Fiddler AI is used by organizations deploying agents across all industries, as well as highly regulated industries like financial services , healthcare , insurance , and government ." Seven trust-gap categories, including insurance, indemnification and dispute process, show no public information for Fiddler's own product. (full piece: `fiddler-ai-governance-vendor-no-insurance-disclosure`) - **Kortix Suna** — an open-source coding agent that gates merges behind an explicit, human-approved permission grant: "Merge is default-deny for an agent and granted explicitly in kortix.yaml — so an agent cannot widen its own reach without a change someone else approves." The agent can be started by cron schedules or signed webhooks with no person present. SOC 2 Type I and Type II audits are in progress, not complete. Four trust-gap categories, including insurance and indemnification, show no public information. (full piece: `kortix-suna-ai-open-source-agent-permission-gating`)

## What changed in the index this week

Real counts, queried August 10, 2026 at 16:22 UTC:

- **Published entities: 1,785** of 2,868 tracked (1,743 AI agents, 42 assurance/supplier vendors); 1,083 remain in the candidate queue. - **31 entities cleared to published status in the seven days ending August 10, 2026** — by publish date: Aug 4 — 17, Aug 5 — 3, Aug 8 — 5, Aug 9 — 6, Aug 10 — 0 as of the query time above. Aug 3, Aug 6 and Aug 7 each recorded zero. - **26 wire articles published to date**, 3 of them dated August 10, 2026 (above). - **Scorecard coverage: 1,169 of 1,785 published entities (65.5%) carry at least one trust-gap dimension; 58 (3.2%) carry the complete five-dimension scorecard.** Judged against the index's own bar for rigor-equivalent coverage — 12 or more present fields *and* a complete five-dimension scorecard — 17 entities qualify, level with the August 9 count. - **86 new scorecard rows were recorded in the same seven days.** Most scored dimensions still come from `v1-auto`, an automated scorer structurally unable to award a literal 0 or 3 on any dimension; only entities scored under the manual `v1` rubric can register the true floor.

## One assurance/insurance development: DeepOpinion automates insurance claims, discloses none of its own coverage

DeepOpinion's marketing names insurance claims processing, loan processing and policy administration as target workflows, and the company describes maintaining "core operations in insurance, finance, and governance that involve complex validation checks, approval workflows, and audit trail maintenance" for those decisions. What its published pages do not state is what happens if one of those regulated, automated decisions is wrong: no named insurer, no coverage limit, no indemnification clause and no dispute process appear anywhere on the site AI Dispatch reviewed.

That gap is not unique to DeepOpinion. Of the published entities scored on the relevant dimensions as of this issue, 61 of 67 (91%) score 0–1 on `insurance_indemnity`, 547 of 871 (63%) on `audit_trail`, and 312 of 400 (78%) on `outcome_settlement` — the three dimensions this index treats as a trust gap at that band. Most of that population is scored by the automated `v1-auto` scorer, which cannot register a literal 0; only manually scored entities can be said to sit at the true floor.

## Methodology

Full rubric and sourcing standard: `/methodology`. Every figure above was queried directly against the live database at draft time (August 10, 2026, 16:22 UTC) rather than carried forward from a prior issue.

---

*AI Dispatch is not a growth-hacked directory. Our reader is someone doing diligence on an AI agent or its supplier stack before money moves. We publish absence as a finding and cite everything. If we ever stop doing that, stop trusting the newsletter.*