The progressive media ecosystem
is embedded in a $1.8 trillion IPO pitch.
The disclosed payment: nothing.

Step 1 — What the supply function is (and isn't)

The progressive media ecosystem is not a large share of Grok's training by volume — it is a rounding error there. The claim, which is structural and documented, is different: the progressive media ecosystem is a load-bearing supplier to Grok's grounding moat — the real-time news-grounding function that lets Grok answer questions about what's happening now. That function is Grok's primary commercial differentiator. Training volume and grounding quality are different things; this registry is about the latter.

Step 2 — Per-entity annual licensing value (comparables method)

Each entity's annual value is drawn from the existing Valuation Registry, which anchors to disclosed AI-licensing deals by revenue tier and citation-band premium:
High tier base · Med tier base · Low tier base
Band premium: Band A (apex) 2.5× · Band B (bridge) 1.6× · Band C / none 1.0×

Step 3 — Cohort aggregate → embedded value → capitalization

Aggregate annual modeled licensing value across all 159 entities: ≈$189M/yr.
AI data asset capitalization multiple: 20–30× (benchmarked against xAI's own $250B pre-merger valuation, which implied a massive multiple on its data dependencies — the ecosystem's 20–30× is conservative by comparison).
Total progressive media ecosystem embedded value: Floor $2.7B · Mid $6.5B · High $30B

Step 4 — Per-entity proportional share

Each entity's share of the embedded value is its weight in the aggregate:
entity_embedded = (entity_annMid / total_annMid) × ecosystem_embedded_total
This is the same proportional logic used in any royalty-pool allocation — the entity's relative contribution to the aggregate determines its share of the collectively negotiated value.

What "layer 3 share" means

Layer 3 = the AI/data sub-entity of SpaceX at IPO (xAI, into which X had already been folded before the SpaceX merger). Mid estimate: ~$325B — anchored to xAI's $250B February-2026 merger valuation (already inclusive of X), adjusted toward the IPO. Entity's Layer 3 share = entity embedded mid / $325B. This is a relative positioning figure, not a legal claim.

Entity

The named outlet, publication, podcast, newsletter, think tank, or individual creator. 159 entities drawn from the Progressive Media Ecosystem Atlas (Issue 02, Phase 8b). Each is an active, identifiable content producer whose public posts appear in X's content stream — and therefore in Grok's real-time grounding pool.

Section

The Atlas category the entity belongs to: Legacy Print, Digital Native, Nonprofit Investigative, Substack Anchors, and so on. Section groupings reflect structural similarities in business model, editorial identity, and platform behavior — not political proximity. Use the section filter above the table to isolate any category.

Band

The entity's citation-centrality class within the ecosystem — how load-bearing it is as a source that other outlets and grounding queries depend on.
A Apex supplier — original reporting that other outlets, AI grounding queries, and downstream aggregators treat as primary source. Highest grounding utility.
B Bridge — significant original reporting with broad cross-platform distribution; cited frequently but not exclusively sourced.
C Chorus — aggregator or commentary layer; amplifies and reframes rather than originating. Grounding utility derives from reach, not sourcing primacy.
n/a No citation-band assignment; typically print-primary or very small digital footprint.

Grounding index

A relative 1–100 score — not dollars — measuring how indispensable an entity's content is to keeping Grok's real-time answers accurate and sourced. Derived from citation band (A/B/C) combined with revenue tier (H/M/L). It drives the band premium in the annual comparable calculation, and it explains why two entities at the same revenue tier can have different modeled values: a Band A mid-tier outlet contributes more grounding signal than a Band C high-revenue aggregator.

Annual comparable

The estimated annual licensing value of this entity's content supply, modeled on disclosed AI-publisher deals: News Corp ≈$50M/yr at the high end (R8); Reddit ≈$60M/yr for real-time access (R9); smaller outlets reportedly pitched at a $1M floor (R10). Each entity's figure is the product of its revenue tier base rate and its citation-band premium. This is what the market has demonstrably paid for comparable content in comparable AI licensing contexts — not what this entity has been offered or received.

SpaceX share

The entity's proportional share of the modeled progressive media ecosystem embedded value within the SpaceX IPO stack. Calculated as: (entity annual comparable ÷ total ecosystem annual comparable) × total ecosystem embedded value. The total embedded value ($2.7B floor / $6.5B mid / $30B high) is the aggregate annual comparable capitalized at the AI data asset multiple implied by xAI's own pre-merger valuation. Use the Conservative / Mid / Bullish toggle to switch scenarios. This figure is illustrative — it is a proportional allocation of a modeled total, not a disclosure from SpaceX or xAI.

Compensation

What appears in the SpaceX S-1 or any publicly disclosed contract as compensation paid by xAI or X to this entity for AI content use. For all 159 entities in this registry as of May 2026: none on record. The SpaceX S-1 is a sworn SEC filing in which material contracts must be disclosed — the Anthropic compute deal ($1.25B/month) appears; no progressive media ecosystem licensing agreement does (S1-E). This column records what is in the public record, not what may or may not exist in private arrangements below SEC materiality thresholds. If a deal were publicly disclosed, this field would update to reflect the terms. The column is descriptive only — it is not an input to any calculation in this registry.

The S-1 — primary source (SpaceX's sworn SEC disclosure)

The S-1 is the registration statement SpaceX filed with the SEC ahead of the IPO. It is a primary source: sworn disclosure, legally accountable. The public version dropped May 20, 2026. It is the strongest receipt in this registry.

• S1-A — SpaceX pitches itself as an AI company: AI language = 47% of segment-specific S-1 text; 93% of stated $28.5T TAM attributed to AI; seven of 12 growth strategies relate to AI: Morningstar S-1 analysis · PitchBook S-1 dissection
• S1-B — AI segment financial reality: 6.7% of revenue, $14B free cash flow loss in 2025; Starlink (61% of revenue) is the only profitable segment: Morningstar
• S1-C — Grok user base: 117M Grok feature users out of 550M MAUs; S-1 counts X platform users as Grok users: Odaily S-1 memo analysis
• S1-D — S-1 discloses $530M litigation reserve for Grok image-generation class actions (Jane Doe v. X.AI Corp. Jan 2026, and two March 2026 cases); S-1 states range of additional loss cannot be estimated: Yahoo Finance S-1 coverage
• S1-E — S-1 does not disclose any content licensing agreements with journalism publishers or content creators. The absence of disclosure in a sworn SEC filing is a documented structural fact.
• S1-F — Anthropic (direct Grok competitor) paying SpaceX $1.25B/month for compute access through May 2029; largest single disclosed AI-segment revenue item: PitchBook

The IPO — valuation and structure

• R-IPO-1 — SpaceX targets ≥$1.8T IPO valuation (lowered from $2T after investor consultations); seeks to raise up to $75B; formal marketing expected June 4, pricing June 11: Yahoo Finance / Bloomberg, May 29 2026
• R-IPO-2 — SpaceX to list on Nasdaq under ticker SPCX; Goldman Sachs, Morgan Stanley, BofA, Citi, JPMorgan leading: Yahoo Finance

The merger — xAI and X absorbed into SpaceX

• R2 — SpaceX acquired xAI in February 2026; xAI valued at $250B, SpaceX at $1T at merger: Britannica / Encyclopedia
• R3 — xAI dissolved into SpaceX (announced May 6–7 2026); Grok and X now sub-entities of the IPO entity; X consolidated into SpaceX's AI segment on the S-1 balance sheet: Odaily S-1 analysis

Grok's competitive differentiation — the grounding claim at source

• R5 — Musk at Grok's launch, November 4 2023: "Grok has real-time access to info via the X platform, which is a massive advantage over other models." Direct quote, primary source: Fortune · TechCrunch
• R5b — Independent corroboration of X firehose as structural competitive advantage; "the only AI lab with embedded access to a major social platform's real-time data and 600+ million monthly users": The AI Rankings
• R5c — S-1 itself characterizes Grok's primary use case as real-time fact-checking via X; Grok described as providing "AI-generated context underneath other users' posts": TechRadar

Default-on ingestion — the platform's own terms

• R6a — July 2024: X activates hidden default-on toggle using public posts to train Grok; buried in privacy settings, activated without user notification: Silicon Republic
• R6b — August 2024: Ireland's DPC launches emergency High Court proceedings against X; first use of DPC emergency powers to stop AI training. September 2024: X agrees to permanently suspend use of EU/EEA users' posts for Grok training: Anonyome (full timeline)
• R6c — November 2024: X updated terms of service remove even the opt-out option for non-EU users; starting November 15 2024, X can use, modify, and transmit all content "in any and all media or distribution methods now known or later developed, for any purpose": Cybernews
• R6d — April 2025: Irish DPC opens formal statutory inquiry into whether X processed personal data in public posts without consent for Grok training; inquiry ongoing at IPO filing: TechRadar

Attribution — the citation studies

• R7a — CJR / Tow Center: eight AI search engines compared; all bad at citing news sources: CJR Tow Center
• R7b — arXiv study evaluating commercial AI chatbots (Grok 4 included) as news intermediaries; measures accuracy on same-day news questions, evidencing reliance on current journalism for real-time grounding: arXiv:2605.22785

Licensing comparables — the valuation floor

• R8 — News Corp–OpenAI: reported up to $250M over 5 years (≈$50M/yr): Variety
• R9 — Reddit–Google: reported ≈$60M/yr for real-time content access: CJR analysis
• R10 — Smaller outlets reportedly pitched at "starting at $1M": A Media Operator
• R11 — Market shift toward usage-based / real-time-fetch pricing; grounding becoming a paid licensing tier: Media and the Machine

Supplier asymmetry — the Atlas foundation

• R1 — Progressive Media Ecosystem Atlas (Issue 02), Phase 8b: 5–15× citation asymmetry finding; 159 active entities; methodology foundation for this registry: Issue 02 · Atlas working data

Litigation posture — ecosystem entities in AI copyright suits

• R12 — CIR (Mother Jones/Reveal) and The Intercept consolidated into NYT v. OpenAI/Microsoft (SDNY): Nieman Lab
• R13 — Raw Story / AlterNet DMCA suit dismissed on standing (not merits): Ronin Legal

Full receipt set for Issue 03 published at openissuessoft0526.netlify.app/issue-03-receipts. S1-E (absence of content licensing disclosure in the SpaceX S-1) is a structural fact verifiable directly in the filing; no URL exists because a non-disclosure cannot be hyperlinked.