{ "issue_id": "SS-OUTLOOK-2026-09-BASELINE", "version": "1.0.1", "status": "published", "author": "Mike Ye", "prepared_on": "2026-09-16", "published_on": "2026-09-16", "evidence_cutoff": "2026-09-16", "period_start": "2026-09-01", "period_end": "2026-09-16", "dataset_version": "1.0.6", "methodology_version": "1.0.0", "intended_canonical_path": "/ai-data-transactions/outlooks/2026-09-baseline/", "title": "The emerging market is pricing permission to learn", "subtitle": "September 2026 opening baseline · four approved observations · not a full-month market census", "thesis": "Our opening sample supports a market for permission to learn from proprietary corporate data. It does not yet establish a general price for corporate operating history. Tempus provides a contractually identified license-fee reference; OpenFold demonstrates a way to learn across protected archives; NAVER and TotalEnergies show strategic co-development. The economic unit to track is the usable learning right, together with its restrictions and related obligations.", "transaction_ids": [ "SS-AIDT-2025-0001", "SS-AIDT-2025-0002", "SS-AIDT-2026-0004", "SS-AIDT-2026-0007" ], "sections": [ { "heading": "What this issue measures", "paragraphs": [ "This opening issue freezes dataset v1.0.6 at 16 September 2026 and includes only human-reviewed observations. It covers four approved transactions across three industry classifications. Five provisional records are excluded from every measure below. These are selected public observations, not estimates of total market activity.", "Two approved announcements fall within September 1–16: NAVER's consortium on September 3 and TotalEnergies–Mistral on September 15. Tempus–Pathos, disclosed April 23, 2025, and the OpenFold expansion announced October 1, 2025, are historical observations admitted to the approved baseline this month. The September 14 OpenFold performance report updates an existing observation; it is not a new deal.", "The first issue has no comparable previous monthly sample. No growth rate, deal-velocity acceleration or market-share claim is calculated. Subsequent monthly issues should show new announcements, historical additions and updates separately." ], "supporting_transaction_ids": [ "SS-AIDT-2025-0001", "SS-AIDT-2025-0002", "SS-AIDT-2026-0004", "SS-AIDT-2026-0007" ] }, { "heading": "Pricing: a license fee, a program commitment, and two undisclosed arrangements", "paragraphs": [ "Tempus's SEC filing identifies $200 million of data-license fees over three years, including $50 million payable upfront. Pathos can settle up to half in preferred stock. Separate $35 million obligations run from AstraZeneca to Tempus and from Tempus to Pathos. We retain the gross contractual license amount while showing the linked obligations; we neither net them into an invented asset price nor add them to the license amount. Payment completion is unverified.", "TotalEnergies discloses more than €100 million over three years for a broader AI program. That lower-bound commitment includes activities beyond data access and does not price the geoscience corpus separately. OpenFold and NAVER provide no standalone cash price in the approved records.", "The USD license commitment and EUR program commitment are separate evidence categories. There is no combined transaction-dollar headline, average dataset price, annualized licensing run rate, or implied corpus multiple. One identified license fee cannot establish a defensible valuation range." ], "supporting_transaction_ids": [ "SS-AIDT-2025-0002", "SS-AIDT-2026-0007", "SS-AIDT-2025-0001", "SS-AIDT-2026-0004" ] }, { "heading": "Learning usefulness and price discovery are different observations", "paragraphs": [ "OpenFold has the highest Asset Score in this sample, 87, but a Transaction Signal Score of 40. Tempus scores 82 and 75 respectively. These analyst assessments illustrate why Strategic Signal keeps two scores: useful data can sit inside an arrangement that reveals little about price, while an explicit contractual payment supplies stronger economic evidence.", "The consortium reports training on 20,167 private structures and evaluation on 1,056 held-out structures. High-quality interface predictions increased from 35.6% to 52.1%; correct ligand poses increased from 28.9% to 46.8%. Those changes are 16.5 and 17.9 percentage points, respectively. This is consortium-reported predictive performance, not independent replication, clinical benefit, or a valuation.", "All four approved records involve continuing businesses or institutions. Three are classified as renewable corpora; Tempus's renewal status is unknown. None establishes a verified standalone ownership sale in this sample. This describes our coverage, not evidence that stranded assets or ownership transactions are absent from the wider market." ], "supporting_transaction_ids": [ "SS-AIDT-2025-0001", "SS-AIDT-2025-0002", "SS-AIDT-2026-0004", "SS-AIDT-2026-0007" ] }, { "heading": "Where we would look next", "paragraphs": [ "Strategic Signal analysis: prioritize industrial manufacturing and equipment maintenance first. Interventions, machine state and subsequent failures or recovery could form economically useful learning histories. Energy and geoscience supply the closest approved operational analogue; this is an inference, not proof of a manufacturing-data transaction.", "Second, examine insurance underwriting, claims and fraud decisions. Outcomes can make records useful for learning, but consent, changing risk regimes and rights usability could prevent an archive from becoming a transferable asset. Our approved sample contains no insurance deal; this remains a research hypothesis.", "Third, watch logistics and aviation maintenance. Repeated decisions linked to downtime, routing and service outcomes may support task-specific models. Fragmented ownership and poor linkage between actions and outcomes are plausible constraints. This ranking expresses research priority, not a quantified probability or industry valuation forecast." ], "supporting_transaction_ids": [ "SS-AIDT-2026-0007", "SS-AIDT-2026-0004" ] }, { "heading": "M&A implications: underwrite a rights package", "paragraphs": [ "Our analysis: the diligence unit should be the corpus plus permitted use, provenance, restrictions, update obligations and model rights. A seller's ownership of a software business does not by itself demonstrate the ability to license employee, customer or third-party material for learning. Those uncertainties can constrain commercial usability even when a corpus is difficult to reproduce.", "For a buyer, the critical question is what an additional corpus enables for a defined model task and whether that improvement survives evaluation outside the seller's examples. The price of a training license can differ materially from the price of an exclusive, transferable archive with refresh rights. Scores support comparison of evidence and strategic usefulness; they are not pricing multiples.", "The next valuable benchmark is a second independently disclosed, separately priced learning right with sufficient scope and settlement detail for comparison. More partnership headlines would enlarge the register without necessarily improving valuation insight." ], "supporting_transaction_ids": [ "SS-AIDT-2025-0002", "SS-AIDT-2025-0001", "SS-AIDT-2026-0007" ] } ], "forecasts": [ { "forecast_id": "SS-FC-2026-001", "version": "1.0.0", "status": "unresolved", "issued_on": "2026-09-16", "prepared_on": "2026-09-16", "evidence_cutoff": "2026-09-16", "horizon_start": "2026-09-17", "horizon_end": "2027-09-16", "intended_canonical_path": "/ai-data-transactions/outlooks/2026-09-baseline/#ss-fc-2026-001", "thesis": "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.", "supporting_transaction_ids": [ "SS-AIDT-2025-0001", "SS-AIDT-2025-0002", "SS-AIDT-2026-0004", "SS-AIDT-2026-0007" ], "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": "moderate", "confidence_rationale": "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.", "outcome_observations": [], "canonical_path": "/ai-data-transactions/outlooks/2026-09-baseline/#ss-fc-2026-001" }, { "forecast_id": "SS-FC-2026-002", "version": "1.0.0", "status": "unresolved", "issued_on": "2026-09-16", "prepared_on": "2026-09-16", "evidence_cutoff": "2026-09-16", "horizon_start": "2026-09-17", "horizon_end": "2027-09-16", "intended_canonical_path": "/ai-data-transactions/outlooks/2026-09-baseline/#ss-fc-2026-002", "thesis": "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.", "supporting_transaction_ids": [ "SS-AIDT-2026-0007" ], "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": "low", "confidence_rationale": "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.", "outcome_observations": [], "canonical_path": "/ai-data-transactions/outlooks/2026-09-baseline/#ss-fc-2026-002" }, { "forecast_id": "SS-FC-2026-003", "version": "1.0.0", "status": "unresolved", "issued_on": "2026-09-16", "prepared_on": "2026-09-16", "evidence_cutoff": "2026-09-16", "horizon_start": "2026-09-17", "horizon_end": "2027-09-16", "intended_canonical_path": "/ai-data-transactions/outlooks/2026-09-baseline/#ss-fc-2026-003", "thesis": "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.", "supporting_transaction_ids": [ "SS-AIDT-2025-0002", "SS-AIDT-2026-0007" ], "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": "low", "confidence_rationale": "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.", "outcome_observations": [], "canonical_path": "/ai-data-transactions/outlooks/2026-09-baseline/#ss-fc-2026-003" } ], "review_protocol": "Publication requires owner review of this exact issue and forecast wording. On approval, record the actual issue date and approval evidence; if the date changes, review the forward horizon and evidence cutoff before issuing. Subsequent monthly reviews append dated outcomes with supporting record IDs. Preserve original forecast text and versions; revisions create a successor version. Status can be supported, contradicted, unresolved or revised. No automatic publication or additional scheduler.", "limitations": [ "Four selected approved observations cannot establish market size, growth, sector shares or a valuation range.", "The two historical 2025 observations are not September 2026 deal flow.", "Public disclosure and research selection favor identifiable counterparties and large programs.", "Contractual commitments are not verified realized cash receipts.", "Unknown terms are not granted rights; analyst scores are not measured investment returns.", "Renewable classification does not establish an enforceable refresh entitlement.", "No prior issued forecasts exist in this opening issue; no hit rate is calculated." ], "canonical_path": "/ai-data-transactions/outlooks/2026-09-baseline/", "publication_review": { "status": "human_reviewed", "reviewer": "Mike Ye", "approval_channel": "owner_conversation", "reviewed_at": "2026-09-16T22:11:57.265Z", "approved_draft_sha256": "357a2caa39095081bca73348eafa2d33f11ec8e79be3381e1d261c3b787f6fd8" } }