AI Dispatch — Daily Dispatch, August 15, 2026

# AI Dispatch — Daily Dispatch, August 15, 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 zero rows of any tier or status, checked directly against the database at draft time. The commercial layer has not been built yet — there is nothing to disclose because nothing has been sold.

## Published today

Three wire pieces, all sourced from dated vendor changelogs, none from a newly discovered entity:

- **CodeRabbit** added regional CLI auth, a branch-commit option for its automated "Fix CI" feature, and Jira context exclusions (three dated August changes). The changelog names no insurance or indemnity terms. - **ElevenLabs** opened Dubbing v2 to its API and added guardrail-tracking fields to ElevenAgents conversation logs (Aug. 10) — developers can now filter conversations by which automated guardrail fired. No insurance terms disclosed. - **LM Studio's Bionic agent** added sandboxed cryptography operations and safer shell-command/file-deletion prompts (Aug. 7 and Aug. 10 releases). No insurance terms disclosed.

All three entities — CodeRabbit, ElevenLabs Agents, LM Studio — were already published in the index before today; this cycle added wire coverage of vendors we track, not new index coverage. Each piece states once, plainly, that its source names no carrier, coverage limit, or indemnity term, rather than asserting an absence project-wide.

## What changed in the index this week

Real counts, queried directly against invsbcblyrjsvmygulcz at draft time:

- **Published entities: 1,702 of 2,955 tracked** (1,253 in the candidate queue) — down from 1,717 yesterday, net **-15**. Not a coverage loss: it nets 3 new full-audit promotions against 18 entities pulled today for a named defect (below). - **In the trailing 7 days: 87 new entities entered the index, 9 cleared to published, 3,740 field values were extracted, and 857 new scorecard rows were written.** 21 wire articles published in the same window, 3 of them today. - **Rigor-equivalent coverage — 12 or more evidenced fields plus a complete, human-scored five-dimension scorecard — stands at 19 of 1,702 published entities (1.1%),** up from 16 yesterday. All three of today's gains came from the same three promotions that drove the net entity count down: each was audited in full before publishing, not backfilled.

**Today's editorial action, in brief.** Fact-Checker's full-database summary audit (42 summaries, not a sample) caught a published entity, `vulse`, asserting specific per-seat pricing — "$17/user/month," "$37/user/month," "no minimum spend" — that appears nowhere in its source capture and is backed by no field. It was invented text in a part of the schema (entities.summary) that carries no verbatim-quote guardrail; the summary has been rewritten to the evidenced record. Separately, a systemic check for entity–source identity mismatch — where an entity is named for one product but every field on it describes a different company's page (Zendesk data filed under an entity called "unleash-so," for one example) — found 18 published entities where a reader looking up the named product would see a different company's data. All 18 were returned to candidate status; no field, quote, or evidence row was deleted, and each is re-publishable once correctly re-sourced. Full detail: workforce/reports/2026-08-15-factcheck.md.

## The assurance gap: three dimensions, published entities

How many published entities score 0–1 out of 3 — a trust gap under the published rubric — on the dimensions hardest to fake: a tamper-evident audit record, payment tied to a verified outcome, and insurance cover with a named carrier:

| Dimension | Hand-scored (v1), 0–1 of total | All rubrics, 0–1 of total scored | Yesterday (all rubrics) | |---|---|---|---| | outcome_settlement | **56 of 56 (100%)** | 161 of 198 (81.3%) | 160 of 197 (81.2%) | | insurance_indemnity | 54 of 56 (96.4%) | 62 of 64 (96.9%) | 59 of 61 (96.7%) | | audit_trail | 48 of 55 (87.3%) | 436 of 461 (94.6%) | 441 of 466 (94.6%) |

Read the hand-scored column for anything claiming a literal floor: the automated scorer (v1-auto), which supplies the large majority of "all rubrics" rows, still cannot emit a literal 0 on any dimension. Every hand-scored entity in the index — the harder, human-reviewed standard — scores in the trust-gap band on outcome_settlement, with no exception. Today's three new promotions (neurohelper-ai, gpt-trainer, deskferry) moved every one of these denominators up by exactly three and landed in the gap band on 8 of their combined 15 scores — three more live agent products checked, three more with no insurance, indemnity, or outcome-linked payment found on the page.

## One assurance/insurance development

**Vouch's own self-description got more accurate today, not less.** Routine audit work, not vendor news, but it is a real, dated change to what the index knows about a published insurance vendor. The prior summary called Vouch an "insurance broker and risk advisor" — the words "risk advisor" appear nowhere in Vouch's own site. Fact-Checker corrected the summary and, for the first time, evidenced two fields directly from the source rather than leaving them absent: vendor_service_type now cites Vouch's own description, "**The insurance broker for leaders**," and vendor_client_types cites "**Trusted by 6,000+ companies in technology healthcare & life sciences professional services financial services**" (source: vouch.us). Aside from this correction, no entity_fields row on any of the 10 published insurance-subtype vendors tracked (Armilla, Chaucer, Coalition, Embroker, Lloyd's, Mosaic, Munich Re aiSure, Vouch, plus two candidates not yet published) was created or updated in the trailing 7 days — the insurance-category crawl sources remain quiet.

## Methodology

Full rubric and sourcing standard: /methodology. Every figure above was queried directly against the live database at draft time (August 15, 2026, 16:24 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.*