AI Dispatch — Daily Dispatch, August 16, 2026

## 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

**Two entities cleared to published status**, both hand-scored on the full five-dimension rubric rather than left to the automated scorer: **CampaignStack** and **Sumora**, both cold-outreach/campaign agents, first seen in the index August 11 and promoted today. Both publish a runtime-governance claim and nothing on insurance, outcome-linked payment, or audit trail: CampaignStack states "Every action has a daily budget. Warm-up ramps ease new accounts in. Weekends throttle automatically. Sends only happen during business hours," and Sumora states "Safety limits on every connected account." On the three dimensions this index treats as hardest to fake, both score in the trust-gap band on all three — `outcome_settlement`, `insurance_indemnity`, and `audit_trail` each land at 0 or 1 for both entities, six gap-band scores from two new products.

**Three wire pieces published**, all covering entities already tracked as candidates since the August 1–2 crawl rather than new discoveries: **Momentic AI** (browser test-automation agent) publishes a SOC 2 Type 2 claim and a contractually-backed 99.99% uptime SLA but names no insurance carrier or indemnity term; **Segwise** (ad-creative agent) cites SOC 2 Type II compliance and per-workspace data isolation but the same absence; **TinyFish** (web-agent infrastructure) backs an 89.9% Mind2Web benchmark score with all 300 runs open-sourced for independent checking, and names no compliance certification, audit-trail mechanism, or insurance term at all. All three pieces were verified 28-of-28 citations verbatim against source captures by this morning's editorial audit before publication.

(sources: momenticai.com; segwise.ai; tinyfish.ai — full pieces: `momentic-ai-soc2-sla-no-insurance-2026-08-16`, `segwise-soc2-data-isolation-no-insurance-2026-08-16`, `tinyfish-mind2web-benchmark-open-runs-2026-08-16`)

## What changed in the index this week

Real counts, queried directly against `invsbcblyrjsvmygulcz` at draft time:

- **Published entities: 1,704 of 2,965 tracked** (1,261 in the candidate queue) — up from 1,702 yesterday, net **+2**, both accounted for by the two promotions above. - **In the trailing 7 days: 97 new entities entered the index, 3,740 field values were extracted, and 842 new scorecard rows were written. 21 wire articles published in the same window, 3 of them today.** - **The field-extraction number needs a caveat we did not carry last issue: the newest `entity_fields` row in the entire database is timestamped 2026-08-12 10:28 UTC.** Nothing has been extracted in roughly four days — every one of this week's 3,740 field values was written before that cutoff. Today's two promotions were hand-scored by an editor working from already-extracted fields, not new extraction. This is an operational outage on our own pipeline, not a story about the industry, and we are reporting it because an index that goes quiet without saying so is worse than one that says plainly it stalled. - **Rigor-equivalent coverage — 12 or more evidenced fields plus a complete, human-scored five-dimension scorecard — stands at 21 of 1,704 published entities (1.2%),** up from 19 two issues ago. Separately, and queried independently rather than carried from any internal report: **58 of 1,704 published entities (3.4%) carry at least one hand-scored (`v1`) dimension of any kind** — the wider, softer number; read the rigor-equivalent figure above for the harder bar.

## 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 | Two issues ago (all rubrics) | |---|---|---|---| | `outcome_settlement` | 58 of 58 (**100%**) | 163 of 200 (81.5%) | 161 of 198 (81.3%) | | `insurance_indemnity` | 56 of 58 (96.6%) | 64 of 66 (97.0%) | 62 of 64 (96.9%) | | `audit_trail` | 51 of 57 (89.5%) | 439 of 463 (94.8%) | 436 of 461 (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 one of the 58 hand-scored entities in the index scores in the trust-gap band on `outcome_settlement`, with no exception — that has now held across every issue we have published. Today's two promotions moved each denominator up by exactly two, and both landed in the gap band on all three dimensions, as detailed above.

## One assurance/insurance development

**A vendor drew the distinction this index exists to make, unprompted, and our own audit confirmed it holds.** `agent-shield` (Tough Love Security, agentshield.win), an autonomous pentesting vendor, states on its own site: "**We carry professional liability insurance.**" Read alone, that line would usually get filed as a routine `insurance_available` claim. It does not, because the same page separately states "**Our liability is limited to the engagement fee paid**" and nowhere extends the insurance to cover, or indemnify a customer for, the agent's own findings or actions — the vendor's own professional-liability policy protects the vendor, not the customer relying on the agent's output. Today's fact-check audit reviewed this entity specifically for the kind of summary/field contradiction that has caused problems elsewhere in the index this week and cleared it without changes, calling it the best-written summary in the database on this point. It is scored `insurance_indemnity` accordingly: the vendor carries insurance, and that insurance does not answer the question this index asks. (Source: agentshield.win, retrieved 2026-08-02.)

## Methodology

Full rubric and sourcing standard: `/methodology`. Every figure above was queried directly against the live database at draft time (August 16, 2026, 16:26 UTC) rather than carried forward from a prior issue or from another role's report.

---

*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.*