Learning rights are becoming more explicit than comparable prices

The approved evidence snapshot

11 approved observations; 5 provisional records excluded. 8 announcements in 2026-09-01–2026-09-30; 3 historical observations. Median Asset Score 84; median Transaction Signal Score 68.5. These describe the sample, not the market. Evidence coverage varies by record.

What changed in the approved sample

This September interim month-end review freezes dataset v1.0.15 with eleven approved observations and five excluded provisional records. The opening issue froze four approved observations in v1.0.6. The seven added approvals represent research coverage expansion, not seven new deals announced after the forecast began.

Five additions are pre-horizon observations: Bolt–Arcee was announced September 14; two distinct Silencio licenses were disclosed September 7; Tempus–Recursion originated in November 2023, with a September 2026 amendment; and Harell’s founder post is dated September 15 and already describes the same named datasets and governed learning arrangement. Harell’s September 23 announcement concerns supporting CoreWeave infrastructure. The corrected registry uses the earliest source-stated disclosure identified, not a verified contract execution date. GuideAI–vRad is a September 28 renewal of an existing arrangement. OneMedNet is a September 30 new arrangement.

Eight approved records carry September announcement dates in the frozen dataset, and three carry earlier dates. That mechanical count includes a renewal and Harell’s source-stated first-disclosure date. It is not a count of eight newly created rights relationships. Evidence is cut off at September 30, 2026, 18:35 UTC, before the calendar month has fully closed.

The new information is in the rights package

OneMedNet reports a multiyear training license, retained source-data rights and participation in downstream derivative-data licensing. Harell describes learning inside a controlled environment with data-owner compensation per training run and model-builder ownership of resulting models. GuideAI reports renewed exclusive model-development access to vRad imaging. These observations show different ways to divide permitted learning, source control and downstream commercialization.

Strategic Signal analysis: diligence should price the enforceable rights package and its restrictions. Separate permission to train, rights in resulting models, rights to commercialize a derivative dataset, and rights to sublicense raw records. An announcement of one permission does not supply the others. Longitudinal clinical data and a large network do not by themselves establish linked interventions and outcomes or Human Response H2/H3.

More economic observations do not yet make comparable prices

The approved sample adds a $250,000 Silencio dataset contract and the Tempus–Recursion amendment's $42 million of scheduled third-through-fifth-anniversary license fees. Both are outside the forecast's new-arrangement window. Recursion's amendment specifies three $14 million installments with at least $4 million of each in cash, while reducing the aggregate unique-record cap and changing other obligations. These are contractual terms, not verified receipts or a like-for-like price change.

OneMedNet’s seven-figure disclosure is economically relevant, but the exact amount, currency and settlement form are not established by the reviewed announcement. The corrected successor record leaves currency unknown in both economics fields and preserves the disclosed seven-figure wording. No numeric range or currency conversion is inferred.

Tempus–Pathos remains a separately identified $200 million contractual license reference with mixed potential settlement and linked obligations. TotalEnergies' more-than-€100-million program remains a bundled commitment. No sum, average dataset price, annualized run rate, valuation band or score-derived asset valuation is calculated.

The September forecasts remain unresolved

All three forecasts retain their September 16, 2026 issue date, original wording and September 17, 2026–September 16, 2027 observation horizon. No hit rate is calculated. Their original confidence labels remain moderate, low and low.

SS-FC-2026-001 has one conservatively countable new classifiable observation, OneMedNet. Harell is excluded under the conservative September 15 first-disclosure treatment; GuideAI is an excluded renewal. This supplies directional support for retained-control learning arrangements but falls below the five-observation threshold. The final majority test is explicitly due at horizon end.

SS-FC-2026-002 has no qualifying approved manufacturing or equipment-maintenance owner arrangements in this snapshot. The horizon remains open and a documented final sector search has not occurred. This is an unresolved observable-deal prediction, not evidence that private industrial learning arrangements do not exist.

SS-FC-2026-003 has no fully qualifying new pricing reference under the original requirement to distinguish amount, currency, consideration scope and payment form. OneMedNet provides partial supporting evidence, but its reviewed announcement does not state currency or settlement form. A precise price and named buyer are not additional requirements of the original forecast; the seven-figure disclosure is coarse, not independently disqualifying. Historical additions, renewal headlines and bundled budgets cannot resolve the forecast.

What would change the next review

Prioritize evidence that changes a decision: dated first disclosures and event identity; explicit industrial learning rights from unrelated owners; and a separately allocated new data license with amount, currency and settlement form. For OneMedNet, seek these economic details without estimating them from the headline. For Harell, retain the distinction between the September 15 source-stated learning arrangement and September 23 infrastructure news; revise only if new primary evidence changes that event identity.

Strategic Signal analysis: the developing evidence favors studying retained-control monetization, but the selected sample cannot establish market prevalence or valuation. Keep the canonical AI Access ontology's rights and capability facets distinct. The strict data-rights view is the comparison set here; broader retrieval or workflow access does not automatically qualify for a training-rights forecast.

Three dated forecasts

These forecasts retain their original issuance dates, observation windows, wording and criteria. Dated assessments below append evidence without reissuing the forecasts. These are Strategic Signal analysis, not disclosed facts; confidence is qualitative.

SS-FC-2026-001 · moderate confidence

The next sufficiently observable cohort will favor learning licenses, controlled access and co-development over outright data ownership transfers.

Mechanism: Continuing owners can monetize or benefit from learning while retaining operational access and protecting confidential source data.

Scope: Newly announced, subsequently human-approved transactions within the horizon; exclude historical backfill and updates to existing deals.

Assumptions: Meaningful training permission can be separated from ownership. Public disclosure will describe enough rights to classify at least five future observations.

Alternative: Distressed archive sales or exclusive dataset acquisitions become the more observable transaction form.

Confidence: All four opening cases fit the mechanism, but the sample is selected and small.

Confirmation: At horizon end, at least five approved new observations have classifiable ownership terms and more than half explicitly use licenses, local learning or controlled access without a disclosed transfer of underlying data ownership.

Invalidation: With at least five classifiable observations, half or fewer meet that test.

Unresolved: Fewer than five observations have sufficiently clear rights. Unspecified ownership is not counted as retained ownership.

SS-FC-2026-001 · 2026-09-30 · unresolved

Interim unresolved. OneMedNet is one conservative eligible new classifiable observation. Exclude Harell under the corrected September 15 source-stated first-disclosure treatment, GuideAI as a renewal, and pre-horizon backfill. Threshold of at least five and horizon-end majority test are not met.

SS-FC-2026-002 · low confidence

At least two unrelated industrial data owners will announce qualifying learning-rights arrangements involving manufacturing or equipment-maintenance histories.

Mechanism: Expensive failures and repeated interventions can provide task-specific learning examples that generic public text does not reproduce.

Scope: Manufacturing process, quality/yield, industrial equipment or maintenance corpora; exclude software-only archives, ordinary deployments and two announcements from the same owner group.

Assumptions: Owners can link records to interventions or outcomes. At least some owners can grant usable learning rights and disclose the arrangement.

Alternative: Learning stays internal or agreements disclose only software deployment, leaving no qualifying observable rights transaction.

Confidence: An adjacent energy precedent supports the logic, but the approved baseline has no direct manufacturing precedent.

Confirmation: By horizon end, two qualifying announcements within the horizon from unrelated owner groups are approved with explicit model-learning use.

Invalidation: After a documented horizon-end sector search, fewer than two qualifying announcements are found. This misses the observable-deal prediction, not proof that private arrangements do not exist.

Unresolved: Evidence or final sector search remains incomplete; no automatic success based on deployment announcements.

SS-FC-2026-002 · 2026-09-30 · unresolved

Interim unresolved. Zero qualifying approved manufacturing or equipment-maintenance arrangements from unrelated owner groups in this snapshot. No horizon-end sector search has occurred; do not infer failure or absence of private deals.

SS-FC-2026-003 · low confidence

At least one newly announced transaction will disclose consideration specifically allocated to proprietary corporate data learning rights, adding a pricing reference beyond Tempus.

Mechanism: Explicit licenses create payment obligations that may appear in filings or transaction announcements even when larger programs bundle services and compute.

Scope: A new owner/buyer arrangement announced within the horizon; disclosed amount, currency, consideration scope and payment form must be distinguishable. No minimum dollar threshold.

Assumptions: Some counterparties disclose the license leg separately. A monetary allocation reflects a data right rather than merely a total program budget.

Alternative: Material rights remain hidden inside co-development, equity or service agreements without separate pricing.

Confidence: One clear historical allocation establishes possibility, not its future frequency or price level.

Confirmation: At least one approved transaction meeting scope is supported by a filing, contract or company disclosure before horizon end.

Invalidation: No such transaction is found after a documented horizon-end pricing search. A bundled headline or pending bid does not satisfy the test.

Unresolved: Source access or final search is incomplete. A subsequently discovered pre-horizon deal is historical backfill and cannot count.

SS-FC-2026-003 · 2026-09-30 · unresolved

Interim unresolved. OneMedNet supplies directional support for separately material data-license consideration but the reviewed release does not state explicit currency or settlement form. Do not add a new exact-price or named-buyer requirement: neither is stated in the original forecast. Silencio and Tempus–Recursion are excluded pre-horizon/historical arrangements.

Limits and follow-up

Selected approved observations do not measure the total market, its growth, sector shares or a valuation range.

Seven new approvals are not seven new forecast-window transactions.

The corrected snapshot distinguishes Harell’s source-stated September 15 disclosure from September 23 infrastructure news and leaves OneMedNet currency unknown. Contract execution and historical page-version timestamps are not independently established. Prior immutable releases remain unchanged.

No horizon-end search has occurred; all three forecast statuses remain unresolved.

Contractual fees and source-described commercial structures do not verify cash receipts, model uplift or enforceable rights beyond disclosed terms.

AI Access ontology v2.0.0 governs interpretation; historical methodology v1.0.0 scores remain limited to qualified records.

This cutoff precedes the end of September 30; later evidence is not included.

Later monthly issues will append dated assessments with evidence. Original forecast wording and issue dates are preserved; substantive revisions receive a successor version.

Sources and reproducibility

Original facts are attributed to the sources below. Classification, interpretation and forecasts are Strategic Signal analysis.

• Apheris: AISB initiative expansion — October 1, 2025
• Apheris: Founding initiative — March 27, 2025
• Apheris: Consortium performance report — September 14, 2026
• Apheris: AISB network governance — current description
• NAVER: NAVER Cloud Consortium to Develop AI Models Tailored for Cybersecurity
• TotalEnergies: TotalEnergies Announces a Partnership with Mistral to Develop Frontier AI Models for Reservoir Exploration and Engineering
• Tempus AI / SEC: Form 8-K filed April 23, 2025, Item 8.01
• Bolt: What is Bolt Forge?
• Advanced Blockchain AG via EQS News: Portfolio company Silencio signs record voice-data contract as AI labs turn to low-resource languages
• Silencio: Silencio AI
• CoreWeave and Harell Data: Harell Data Selects CoreWeave to Power Its Secure Platform for AI Model Training
• Harell Data: Data Partners
• Harell Data: Why we started Harell Data
• GeekWire: Adaptive Biotech co-founder raises $15M for new startup to rethink how AI trains on science
• Recursion Pharmaceuticals / SEC: September 15, 2026 Form 8-K — Tempus Master Agreement Amendment
• Recursion Pharmaceuticals / SEC: November 9, 2023 Form 8-K — Tempus Master Agreement
• Recursion Pharmaceuticals / SEC: Tempus Master Agreement Exhibit
• Recursion Pharmaceuticals: Recursion Announces Data Collaboration Deal with Tempus
• GuideAI Health: GuideAI Health Corp. and vRad Renew Data Exclusivity Agreement to Develop AI Models
• OneMedNet: OneMedNet Signs Seven Figure Multiyear Agreement to Power a Partner Sold Derivative Data Solution