Archived dispatch

What does "What Stripe data shows about fraud at AI startups" reveal about payments?

Lowconfidence1 sub-claim remains below the evidence threshold

9/21/2026, 2:59:07 PM · llm:mimo:mimo-v2.5 + llm:deepseek:deepseek-v4-flash on 1 step

The dispatch, itemised.

§ IThe decision$0.02 / $0.04
50%$0.02 under cap
Decompose

Breaking down: "What does "What Stripe data shows about fraud at AI startups" reveal about payments?"

Decompose

Identified 2 research target(s) to investigate; these are not established facts

Decompose

Deep mode: up to 4 paid/cached reads plus one bounded gap-expansion pass when needed.

Discover

Discovered 21 verified source(s)

Discover

Recalled 60 past runs on this subject — how these sources performed when they were available.

Discover

ERC-8004 reputation loaded — composite scores on this subject.

Pre-check

Claim-aware portfolio selected 1/1 positive proposal(s): 1 cached + 0 fresh, predicting 2/2 claim(s) above the evidence floor with $0.000000/$0.020000 fetch USDC reserved.

Pre-check

Free-preview pre-check maps an actionable source to every sub-claim (2/2); paid reading may proceed within the budget.

DecideCACHE
Stripe Blog — What Stripe data shows about fraud at AI startups$0.002 · EV 95%

Exact match for the question: the Stripe Blog preview directly states AI companies faced 4.3x more fraud attempts than startups overall in Q3 2025, which answers subClaim 0 (fraud trends at AI startups) and frames subClaim 1 (payments landscape for AI startups). Already cached at $0.002, and Stripe Blog has the best reputation among relevant sources (47/100, 7 citations). — selected for the claim-aware evidence portfolio (targets claims 1, 2; 0 fetch USDC, 1 attention slot).

DecideSKIP
Decrypt — What Stripe's $7 Billion OpenRouter Deal Actually Means for AI$0.002 · EV 30%

Decrypt's Stripe/OpenRouter piece is topically adjacent (Stripe + AI payments) and could touch the payments landscape (subClaim 1), but it is not cached, its abstract is only 126 bytes, and Decrypt has weak reputation here (8/100). Not worth the toll when the exact Stripe source is already cached.

DecideSKIP
CoinDesk: Bitcoin, Ethereum, Crypto News and Price Data — Crypto’s next billion users might be AI agents, and they’re paying with stablecoins$0.002 · EV 25%

CoinDesk covers AI agents paying with stablecoins, loosely relevant to the AI-startup payments landscape (subClaim 1), but it says nothing about fraud findings (subClaim 0) and CoinDesk has the lowest reputation on this subject (4/100). Redundant with stronger cached sources.

DecideSKIP
Stablecoin Ledger — Why USDC settles instantly onchain$0.003 · EV 20%

Stablecoin Ledger explains USDC onchain settlement, which is background payments infrastructure but does not address Stripe's fraud data at AI startups (subClaims 0-1). Weak reputation (21/100) and not directly on-topic.

DecideSKIP
Agent Economy Weekly — x402 turns HTTP 402 into an agent payment rail$0.004 · EV 20%

Agent Economy Weekly's x402 rail is about agent payment mechanics, not fraud trends at AI startups. Tangential to the payments landscape (subClaim 1) but does not answer the core fraud question.

DecideSKIP
Onchain Micropayments Digest — Nanopayments and the $0.000001 floor$0.005 · EV 15%

Nanopayment floor mechanics are unrelated to Stripe's fraud findings at AI startups; no subClaim is supported by the preview.

DecideSKIP
Distributed Systems Notes — Idempotency keys prevent double-spends$0.003 · EV 15%

Idempotency keys prevent double-spends, a payments reliability topic, but it does not speak to fraud rates at AI startups (subClaims 0-1).

DecideSKIP
Garden & Soil Monthly — Building a no-dig raised bed$0.002 · EV 2%

Gardening content, entirely off-topic for payments fraud at AI startups.

DecideSKIP
Retro Game Hardware — Recapping a 1990s console$0.002 · EV 2%

Retro console repair, entirely off-topic for payments fraud at AI startups.

DecideSKIP
Ethereum Foundation Blog — The triage is the product: running AI agents against Ethereum's protocol code$0.002 · EV 10%

Ethereum Foundation post on AI agents triaging protocol code is about security research workflows, not payments fraud at AI startups.

DecideSKIP
Cointelegraph.com News — New York town weighs crypto mining and AI data center ban$0.002 · EV 10%

Crypto mining/data-center moratorium news is unrelated to Stripe fraud data or AI-startup payments.

DecideSKIP
Latent.Space — Ontologies Are So Back: Why AI Agents Are Reviving the Semantic Web$0.004 · EV 10%

Latent.Space on ontologies and semantic web for agents is about agent architecture, not payments fraud.

DecideSKIP
Simon Willison's Weblog — Feeling sad about AI$0.003 · EV 5%

Metadata-only entry with no preview content; cannot establish relevance to fraud at AI startups, and it is not cached.

DecideSKIP
Hugging Face - Blog — TutorMoments: Do AI tutors know when to help and when to hold back?$0.003 · EV 5%

Metadata-only AI tutor post, off-topic for payments fraud.

DecideSKIP
Vitalik Buterin's website — Low-risk defi can be for Ethereum what search was for Google$0.004 · EV 5%

Metadata-only Vitalik post on low-risk DeFi; no preview and not about Stripe fraud data at AI startups.

DecideSKIP
The Coinbase Blog - Medium — In response to the Wall Street Journal$0.003 · EV 10%

Coinbase's 2022 response to the WSJ concerns proprietary trading allegations, not AI-startup fraud or the payments landscape described in the question.

DecideSKIP
Inner Axiom — The Codex — The Esoteric Teachings of the Planet Isis$0.002 · EV 2%

Esoteric astrology content, entirely off-topic.

DecideSKIP
Conzit Labs — Musk Unveils SpaceX's Ambitious AI Data Center Strategy$0.002 · EV 5%

SpaceX AI data-center strategy is unrelated to payments fraud at AI startups.

DecideSKIP
Arc Settlement Benchmarks — Measuring x402 settlement latency on Arc$0.003 · EV 15%

Arc x402 settlement latency benchmarks concern payment finality, not Stripe's fraud findings at AI startups.

DecideSKIP
Web Payments Review — How long do x402 payments take to finalize?$0.002 · EV 15%

Web Payments Review on x402 finalization timing is about settlement speed, not fraud trends at AI startups; weak reputation (7/100).

DecideSKIP
Keryx Engineering (first-party) — Recovering a Keryx paid research job$0.002 · EV 10%

First-party Keryx engineering notes on buyer recovery are about the research-payment workflow, not Stripe's fraud data at AI startups.

Fetch

Agent spend wallet ready: 0x29028Fe1122E17Fe7863A22701e863FE4DaE1aFB (balance sufficient)

Fetch

Reused cached Stripe Blog — What Stripe data shows about fraud at AI startups (free) — S1

Re-evaluate

Sub-claim "What findings does the Stripe data reveal about fraud trends…": 50% covered by S1 — The abstract directly states that AI companies faced 4.3x more fraud attempts than startups overall in Q3 2025, which is a key finding about fraud trends. However, the question asks for 'fraud trends' (plural) and the full article might reveal more trends (e.g., customer abuse patterns mentioned in the abstract but not detailed in the excerpt). The coverage is partial.

Re-evaluate

Sub-claim "How do these findings relate to the payments landscape for A…": 30% covered by S1 — The abstract mentions 'payments' only in the context of Stripe as a platform. The question asks for how the fraud findings relate to the broader 'payments landscape' for AI startups. The excerpt does not provide this connection; it only states the fraud rate. The coverage is minimal.

Re-evaluate

The primary source (S1) is only an abstract, which provides limited findings and no context for the payments landscape. The Decrypt article about Stripe's OpenRouter deal could offer insight into Stripe's strategy in AI payments, which might relate to fraud trends. The CoinDesk article about AI agents paying with stablecoins could provide broader context on the AI payments landscape. Both are low-cost (0.002 each) and fit within the remaining budget of 0.02. Other skipped sources are either off-topic or too specific (e.g., x402 technical details) for the broad question about the payments landscape.

Re-evaluate

Filling gap — buying Decrypt — What Stripe's $7 Billion OpenRouter Deal Actually Means for AI ($0.002)…

Re-evaluate

Paid $0.002 to Decrypt — What Stripe's $7 Billion OpenRouter Deal Actually Means for AI (settled 59f310c9-e…) — S2

Re-evaluate

Filling gap — buying CoinDesk: Bitcoin, Ethereum, Crypto News and Price Data — Crypto’s next billion users might be AI agents, and they’re paying with stablecoins ($0.002)…

Re-evaluate

Paid $0.002 to CoinDesk: Bitcoin, Ethereum, Crypto News and Price Data — Crypto’s next billion users might be AI agents, and they’re paying with stablecoins (settled e660f9b5-7…) — S3

Sufficiency

Final check — "What findings does the Stripe data reveal about fraud trends…": 100% assessed by S1

Sufficiency

Final check — "How do these findings relate to the payments landscape for A…": 0% assessed

Sufficiency

Final coverage assessment — The article 'What Stripe data shows about fraud at AI startups' (S1) directly answers the first sub-claim about fraud findings. However, it does not address how these findings relate to the broader payments landscape for AI startups. The other sources (S2, S3) discuss AI payments but do not connect Stripe's fraud data to the payments landscape. Therefore, only one sub-claim is fully covered, and the second sub-claim remains missing. The assessment does not establish a complete supported answer for every requested part.

Synthesize

Synthesizing a grounded answer from 3 source(s)…

Evidence

Relevance review returned; only checked excerpts can retain support, and review cannot raise it.

Evidence

Verified — S1 supports claim 1 at 90%: “We analyzed attempted fraud rates and customer abuse patterns on Stripe over the past year and found that AI companies faced 4.3x more fraud…”

Synthesize

Drafted answer citing 1 source(s)

Verdict

Confidence: Low — 1 sub-claim remains below the evidence threshold.

Attribute

Stripe Blog contributed 100% → reward $0.02

Settle

Settled $0.02 citation reward → Stripe Blog (13c3aa52-6…)

Done

Done. Spent $0.024 across 3 confirmed/simulated payment(s) to creators.

§ IIIThe settlementweighted · USDC on Arc
100%
1

Stripe Blog

batched

100%$0.02
§ IIThe reading1 cited
Lowconfidence1 sub-claim remains below the evidence thresholddeep researchpreview plan 2/2 claimsportfolio 1/1 · evidence 100%

> ⚠ Low confidence — 1 sub-claim remains below the evidence threshold within budget. Treat this as provisional.

The Stripe data reveals that AI companies faced 4.3x more fraud attempts than startups overall in Q3 2025, based on an analysis of attempted fraud rates and customer abuse patterns on Stripe over the past year . This indicates a significantly higher fraud risk for AI startups compared to the general startup landscape. However, the provided sources do not contain information on how these specific findings relate to the broader payments landscape for AI startups.

Evidence ledger — quotes verified before rewards

  1. What findings does the Stripe data reveal about fraud trends at AI startups?

    90%
    We analyzed attempted fraud rates and customer abuse patterns on Stripe over the past year and found that AI companies faced 4.3x more fraud attempts than startups overall in Q3 2025. [S1] What Stripe data shows about fraud at AI startups
  2. How do these findings relate to the payments landscape for AI startups?

    0%

    No reward-qualifying evidence

Footnotes — each one pays its author

Helpful?
Spent$0.024
To creators100%
Decisions0 bought · 1 cached · 20 skipped
llm:mimo:mimo-v2.5 + llm:deepseek:deepseek-v4-flash on 1 step

Portable research receipt

Take the evidence trail with you

One deterministic JSON bundle binds the answer, visible decisions, exact article versions, claim evidence and a Circle-settlement snapshot under SHA-256. Retain the digest to detect later changes; the self-check is not a publisher or Keryx signature.

Exact receipt still current

1 exact cited article version still match Keryx's current index. The source cited here has published nothing new since this dispatch settled.

Inspect machine-readable audit

Ask a follow-upNew dispatch · creators paid again

Carries this dispatch’s question as context — never its answer. The next dispatch is read from sources bought for it.

From the archive

Related dispatches