X Corp

Legal name. X Corp.

Former and predecessor names. Odeo, Inc. (2005–2007, the podcasting company from whose internal hack-week project Twitter emerged); Obvious Corporation (interim holding vehicle, 2006–2007); Twitter, Inc. (2007–2023); X Corp. (2023–present). The renaming was not a cosmetic rebrand: Twitter, Inc., a Delaware corporation, was merged out of existence into X Corp., a Nevada corporation, with the merger disclosed in an April 2023 filing in the Loomis v. Twitter litigation in California state court. Twitter, Inc. no longer exists as a legal person. The consumer brand migrated from "Twitter" to "X" on 23 July 2023, and the twitter.com domain was redirected to x.com in May 2024.

Founding and incorporation. The Twitter service launched 21 March 2006 (first post by Jack Dorsey). Twitter, Inc. was incorporated in Delaware in April 2007. X Corp. was incorporated in Nevada on 9 March 2023. The parallel holding entity X Holdings Corp. was incorporated in Nevada in 2023 as successor to the Delaware X Holdings vehicles used for the 2022 buyout.

Current ownership chain (as at 14 August 2026).

This chain is not a matter of inference: the European Commission's 5 December 2025 DSA non-compliance decision was formally addressed to X Internet Unlimited Company (XIUC), X Holdings Corp., X.AI Holdings Corp., and Elon Musk personally, on the reasoning that Musk exercises decisive influence and effective control over each entity in the chain.

Incorporation jurisdiction. Nevada (X Corp., X Holdings Corp.). The EU main establishment, X Internet Unlimited Company, is an Irish company registered in Dublin. Ultimate parent SpaceX was incorporated in Delaware in 2002 and has since redomesticated to Texas.

Headquarters. Bastrop, Texas, United States. X vacated the 1355 Market Street, San Francisco headquarters that Twitter had occupied since 2012; Musk announced the Texas relocation in July 2024 and the move was substantially complete through 2025. Ultimate parent SpaceX is headquartered at Starbase, Texas.

Stock exchanges and tickers. X Corp. has no listed equity and no ticker. It has not had one since Twitter, Inc. was delisted from the New York Stock Exchange (ticker TWTR) on 28 October 2022. Public-market exposure to X Corp. is available only indirectly through Space Exploration Technologies Corp., listed on Nasdaq under SPCX since 12 June 2026.

ISIN/CUSIP. Twitter, Inc.'s legacy identifiers (CUSIP 90184L102; ISIN US90184L1026) were retired at delisting. X Corp. has no ISIN or CUSIP for equity. Its senior secured term loan due 2029 trades in the institutional loan market; the specific CUSIP is not publicly disclosed in a form verifiable here.

Fiscal year end. 31 December.

Employees. Twitter, Inc. reported approximately 7,500 employees at the end of FY2021. Musk reduced headcount by roughly 80% within a year of the acquisition. X Corp. was reported at approximately 1,500 employees in August 2023 and approximately 2,500 in 2024 per company statements. Neither X Corp.'s current standalone headcount nor an X-specific figure inside the SpaceX S-1 is publicly disclosed. SpaceX group headcount and the xAI-segment headcount are disclosed only at group level in the S-1.

Sources: Twitter 10-K FY2021 (Tier A); company statements and press reporting FY2022–FY2024 (Tier C/D). The FY2022 and FY2024 figures are company-asserted rather than audited and should be treated as indicative.

Auditor. Twitter, Inc.'s auditor was PricewaterhouseCoopers LLP (San Francisco). X Corp.'s standalone auditor since 2022 is not publicly disclosed. The auditor of the consolidated group whose financial statements now contain X is disclosed in the SpaceX S-1.

Market capitalisation. Not applicable — no listed equity. Reference valuation points for the X business alone:

Website. about.x.com (corporate); x.com (product).

Positioning statement (150 words). X Corp. is no longer usefully described as a social-media company. It is the consumer distribution layer and real-time data supply for an AI infrastructure conglomerate. Since the March 2025 absorption into xAI and the February 2026 absorption into SpaceX, X has ceased to be a standalone economic unit with its own capital structure, its own cost of capital, or its own reporting obligations. Its advertising business — once approximately 90% of Twitter's revenue and USD 4.51bn at its FY2021 peak — was USD 1.8bn in FY2025 and is now a minority contributor inside a segment whose growth is driven by compute leasing and AI subscriptions. What X supplies to the group is strategically distinct from what it earns: approximately 550 million monthly users, approximately 350 million posts per day of fresh human-generated text, and a distribution surface for Grok, X Money and X Chat. Whether that strategic contribution justifies the USD 44bn paid for it in 2022 remains the central unresolved question in the file.

2.1 What the company does

X Corp. owns and operates X, a global real-time public conversation platform combining short-form text posting, image and video sharing, long-form articles, live audio and video broadcasting, direct and encrypted messaging, creator monetisation, and — since 2024 — deep integration with the Grok family of large language models. Since June 2026 it also operates X Money, a consumer financial-services product delivered through subsidiary X Payments LLC in partnership with a chartered bank.

2.2 The company's own characterisation

The SpaceX S-1 filed 20 May 2026 describes the AI segment as a vertically integrated AI platform spanning a frontier LLM (Grok), AI solutions for consumer and enterprise customers, X as a real-time information, entertainment and free-speech platform, and AI computational infrastructure. On X specifically the filing states an intention to drive revenue growth by increasing engagement across users, increasing X Premium subscriber conversion, growing advertising revenue per user, and diversifying the advertising base. It further describes the planned evolution of X into an "everything app" integrating real-time information, communications, media, payments and banking within one consumer app experience.

Critically, the filing frames X primarily as a data asset for Grok rather than as an advertising business: it argues that Grok represents a differentiated approach to AI grounded in truth-seeking and powered by continuous proprietary access to real-time data inflows through its X integration, and that approximately 350 million daily posts give Grok freshness, relevance and contextual awareness that constitutes a competitive differentiator.

2.3 Independent characterisation

The S-1's framing is directionally honest but strategically self-serving in two respects. First, it recasts an underperforming advertising asset as a data asset, which conveniently converts a revenue shortfall into a strategic input whose value need never be marked to market. Second, it does not confront the quality question: if the platform's content is materially contaminated by automated accounts, coordinated inauthentic behaviour and AI-generated output — a risk X's own 2026 anti-spam measures implicitly concede — then the training-data differentiator degrades in the same motion.

The more defensible independent read is that X today performs three distinct economic functions inside SpaceX:

1. A declining but cash-generative attention business. Approximately USD 1.8bn of FY2025 advertising revenue at high incremental margin, against a heavily reduced cost base.
2. A subscription funnel. 4.4 million X Premium/Premium+ subscribers as at March 2026, and the primary acquisition channel for Grok's 1.9 million paid subscribers.
3. A proprietary real-time corpus and distribution surface. The strategic rationale that survived two mergers.

2.4 Revenue model

2.5 Value chain position and customers

X occupies the aggregation layer: it does not create content, does not own distribution hardware, and following the merger no longer owns its own compute. It sits between end users (who supply content and attention at zero marginal cost) and three paying customer types: brand and performance advertisers; individual consumers buying subscriptions and, now, deposit accounts; and enterprise data/API licensees. End markets served are digital advertising, consumer subscription software, enterprise data licensing and — as of 2026 — consumer fintech.

10.1 Stated strategy

The S-1 articulates four strategic pillars touching X directly: increase engagement across users; increase X Premium subscriber conversion; grow advertising revenue per user; and diversify the advertiser base. The advertising sub-strategy is stated as improving performance advertising capabilities, embedding AI to optimise campaigns, and launching richer ad formats that increase advertiser return on ad spend.

The group-level strategy is considerably more ambitious and effectively subordinates X: the filing states that the company believes its next trillion-dollar market is AI compute, that it expects to leverage its rockets and satellites for massive orbital deployments of AI infrastructure, and that this infrastructure will let it develop and monetise Grok faster than AI companies dependent on finite terrestrial power.

10.2 Announced initiatives, last 24 months

10.3 Medium-term targets

X Corp. issues no guidance. Parent-level guidance from the Q2 2026 earnings call: an annualised revenue run-rate of approximately USD 100bn by end-2026, and an internal USD 1tn revenue forecast pulled forward to 2030 from 2031. No X-specific target has been published.

For reference, the 2022 acquisition pitch deck circulated to co-investors projected USD 26.4bn of 2028 revenue — USD 12bn advertising, USD 10bn subscriptions, balance data licensing — and 69 million paid subscribers by 2025 rising to 159 million by 2028. Measured against the S-1: FY2025 advertising of USD 1.8bn is approximately 15% of the 2028 advertising target three years in; 4.4m X Premium subscribers as at March 2026 is approximately 6.4% of the 2025 subscriber projection.

5.1 Core platform — X

X (x.com; iOS, Android, web). The core real-time public posting platform. Text posts up to 280 characters for free accounts and up to 25,000 characters for Premium subscribers; image, GIF, and video attachment; reposts, quote posts, replies, bookmarks, lists, Communities, and Spaces. Target customer: general consumer, with disproportionate concentration of journalists, politicians, finance and technology professionals, and sports and entertainment audiences. Pricing: free, ad-supported. Scale as at March 2026 per the S-1: approximately 550 million users (up from 520 million in December 2025) generating approximately 350 million posts per day.

Note on the user metric. The S-1's approximately 550 million figure is materially below the 600 million MAU X publicly claimed through 2024–2025, and the filing does not reconcile the two. Third-party measurement diverges further: Similarweb put X's worldwide mobile app MAU at approximately 302 million in June 2026, with mobile DAU of 123.7 million in July 2026 (down 7% YoY). These are different denominators — audited total users versus measured mobile-app actives — and should not be reconciled naively, but the direction of travel in the third-party series is unambiguously negative.

"For You" recommendation algorithm. Rebuilt on Grok and open-sourced under Apache-2.0 on 20 January 2026, an unusual disclosure for a ranking system of this scale. In July 2026 X acknowledged that missing data had been affecting the algorithm and adjusted it to surface more content from followed accounts.

Community Notes. Crowd-sourced contextual annotation system, inherited from Twitter's Birdwatch pilot. AI-generated Community Notes entered testing in July 2025. Frequently cited by Yaccarino and Musk as the platform's signature integrity product.

Spaces. Live audio rooms. Live / Creator Studio. Livestreaming, upgraded through 2026 with a livestream command centre for scheduling and managing broadcasts, desktop support, and multi-language captioning (July 2026). Articles. Long-form publishing for Premium subscribers, launched 14 January 2026.

About This Account. Transparency feature launched under Nikita Bier surfacing an account's actual country of origin; materially disrupted networks of accounts misrepresenting their location.

5.2 Subscriptions

Table 5A — Subscription tier stack (US list pricing as at the S-1 filing date)

Table 5B — Paid subscriber base (S-1, as at 31 March 2026)

The tier-level split within each bucket is not publicly disclosed. The 0.8% paid-conversion rate is the single most important operating statistic in the file for anyone assessing the "everything app" thesis: after three years of aggressive feature-gating, fewer than one in a hundred X users pays for the platform.

5.3 Advertising products

X Ads Manager. Fully rebuilt and re-released April 2026, repositioned toward small and medium-sized businesses in the manner of Snap and Reddit self-serve tooling. The S-1 attributes part of the Q1 2026 advertising decline to disruption during the rebuild. Promoted posts, trend takeovers, Amplify pre-roll, and vertical video ads constitute the inventory set. AI-driven performance tools are cited in the filing as the forward differentiator. The S-1 explicitly states that Grok API access is not currently included in advertising rates and that advertisers cannot currently place ads on the Grok API — phrasing that leaves the door open.

Ad-adjacent DSP pilot. In 2025 X piloted prompt-relevance-driven placements with reported CPMs of USD 15–60 and a USD 50,000 minimum spend.

5.4 Creator monetisation

Creator Revenue Sharing (2023–2026). Ad-revenue share against replies from verified users. Closed to new enrolments 8 August 2026; existing members earn through 7 September 2026, with final payments around 11 September. Original Content Rewards Program (from 8 September 2026). Successor programme explicitly designed to reward original reporting, expertise, storytelling and entertainment, and to stop rewarding the practice of adding a text caption to someone else's content. Engagement-bait solicitation is grounds for removal.

5.5 Messaging

X Chat. Launched as a web service December 2025; standalone iOS app launched 28 April 2026 following TestFlight beta from 3 March 2026 (1,000 testers, expanded to 5,000). End-to-end encryption with a device-held PIN-protected key pair; screenshot blocking; message editing; disappearing messages; large file transfer; group chats supporting several hundred participants; no ads and no activity tracking per the App Store listing. Grok integration via long-press "Ask Grok" operates on an unencrypted copy of the selected message — a design compromise that drew immediate privacy criticism. Android version pending.

5.6 Payments and financial services

X Money (X Payments LLC). Beta November 2025; limited release to a subset of US Premium+ users 25–26 June 2026; full US rollout to Premium and Premium+ subscribers 27 July 2026. Head of X Money: Dhruv Batura.

5.7 Grok (xAI product, distributed through X)

Grok is an xAI rather than X Corp. product but is inseparable from X's product surface and subscription economics. Launched November 2023 in beta for Premium+; free to all X users from December 2024; standalone mobile apps early 2025; Grok 4 July 2025; Grok 4.1 and 4.1 Thinking, Grok 4.3, and the Grok 4.20 line through 2026. Grok Imagine provides image and video generation with Remix and Upscale settings and AI companions. Grok MAU: approximately 117 million as at March 2026 per the S-1 — implying only about 21% of X users engage with Grok.

Grok "Edit Image" (December 2025). The feature at the centre of the company's most serious ongoing regulatory exposure. It permitted users to AI-edit images of other users, initially without consent controls or opt-out, and was rapidly used at scale to generate sexualised images of real people, including minors. Image creation on X was restricted on 9 January 2026 following backlash. This is addressed in Sections 14, 18 and 19.

5.8 Data and developer products

Tiered API access (free, basic, pro, enterprise) and firehose-equivalent enterprise data licensing. Data licensing revenue is described in the S-1 as having grown consistently, with an FY2025 incremental contribution of approximately USD 88m reported from the filing. Absolute data-licensing revenue for X in FY2025 is not publicly disclosed; the last audited figure is Twitter's FY2021 USD 0.57bn.

6.1 Preliminary: what can and cannot be built

A conventional five-year financial model of X Corp. cannot be constructed from public sources and is not attempted here. FY2022 through FY2024 contain no audited disclosure at all; FY2025 exists only as a partially disaggregated line inside a combined AI segment. Any dossier presenting a complete five-year P&L, balance sheet and cash flow statement for X Corp. is presenting fabrication. What follows is every audited, regulatory and credibly reported datum, arranged so the gaps are visible.

6.2 Income statement

Table 6A — X / Twitter revenue by line (USD m)

FY2021 is Tier A (Twitter 10-K). FY2025 advertising of USD 1.8bn is Tier A (SpaceX S-1 as reported by Social Media Today). The FY2024 figure of USD 1,685m is derived, not disclosed: the S-1 states advertising revenue grew by USD 115m in FY2025 after declining USD 595m in FY2024, which implies FY2024 of approximately USD 1,685m and FY2023 of approximately USD 2,280m. These derivations are arithmetic from disclosed deltas and are flagged as such.

Table 6B — Disclosed year-over-year revenue movements (USD m)

A material source conflict requiring flagging. Coverage of the S-1 diverges on the direction of FY2025 advertising. CNBC's live coverage of the filing (20 May 2026) reports that full-year 2025 advertising revenue grew by USD 115m after falling USD 595m in 2024, with the USD 100m decline attaching to Q1 2026. Several secondary outlets — including a widely circulated Motley Fool piece and analyses derived from it — present the USD 100m decline as a full-year FY2025 figure. These cannot both be right. The CNBC reading is internally consistent with the S-1's attribution of the Q1 2026 softness to the Ads Manager rebuild, and is the reading adopted here; users should note that a meaningful body of published analysis assumes the contrary.

A second scope conflict: TechCrunch reported an AI-segment advertising line of approximately USD 116m, against Social Media Today's USD 1.8bn X-platform figure. These are almost certainly different carve-outs of the same business — the smaller number appearing to be a Grok-adjacent sub-line — and must not be added. EMARKETER's pre-filing forecast of USD 2.26bn global X ad revenue for 2025, widely cited through 2025, is superseded by the audited USD 1.8bn.

Table 6C — Profitability (USD m)

FY2024 EBITDA of approximately USD 1.5bn and interest expense above USD 1.3bn are Tier C — lender-materials figures reported by Bloomberg in April 2025, presented by management to investors, and not independently audited. The Q3 2025 net loss of USD 577.4m is Tier C via Bloomberg. FY2021 net loss is Tier A.

Table 6D — Reported quarterly revenue, X standalone (USD m, Tier C)

Reported via Bloomberg in December 2025; nine-month 2025 revenue exceeded USD 2bn, with Q3 up 17% YoY. Q1 is derived from the reported statement that Q2's USD 707m was a 2.2% sequential decline.

Gross profit, operating income, pre-tax income, EPS, dividends per share, and all margin ratios for X Corp. are not publicly disclosed for FY2022–FY2025. X Corp. has never paid a dividend. Twitter, Inc. never paid a dividend.

Table 6E — Consolidated parent (SpaceX, recast to include X and xAI) (USD m)

Q2 2026 net loss is reported at USD 541m by one wire source and at approximately USD 1.26bn by another; the discrepancy likely reflects attributable-versus-total or continuing-operations definitions and is unresolved at the time of writing. Note also that Motley Fool coverage transposed the FY2024 profit and FY2025 loss; the sequencing used here (profit in FY2024, loss in FY2025 post-recast) is the internally consistent reading.

6.3 Balance sheet and capital structure

Table 6F — X Corp. capital structure (USD m)

Cash of approximately USD 120–320m through the year to January 2025, rising to approximately USD 1.1bn by April 2025, is Tier C via Bloomberg. Total equity, goodwill and intangibles, working capital, and net debt for X Corp. standalone are not publicly disclosed; goodwill in particular would be highly material given a USD 44bn purchase price against a business generating approximately USD 2.5–3bn of revenue, and its non-disclosure is a significant analytical gap.

Table 6G — Debt syndication and refinancing chronology

Syndicate members: Morgan Stanley (lead and largest holder), Bank of America, Barclays, Mitsubishi UFJ, BNP Paribas, Mizuho, Société Générale. Buyers of the distributed paper included Pimco and Citadel. Post-distribution secondary marks were 98–99c. Group total debt at SpaceX level was USD 29.1bn as at 31 March 2026.

6.4 Ratios

Return on equity, return on assets, ROIC, current ratio, debt/equity, net debt/EBITDA, interest coverage, asset turnover and cash conversion cycle cannot be computed for X Corp. because no balance sheet has been published since FY2021. Two partial ratios are computable from Tier C data and are informative:

An interest cover of roughly 1.15x on approximately USD 1.5bn of EBITDA against more than USD 1.3bn of interest is the arithmetic explanation for why the 2025 refinancing mattered so much, why the equity round and the xAI absorption followed within months, and why the business could not have serviced its capital structure independently for much longer.

6.5 Commentary on trends, inflections and drivers

Inflection 1 — the advertising collapse (2022–2024). Advertising fell from USD 4.51bn in FY2021 to roughly USD 1.69bn in FY2024, a decline of approximately 63%. The USD 595m single-year decline in FY2024 is the largest disclosed annual drop. Drivers were brand-safety withdrawal following moderation changes, the loss of enterprise sales capability after the workforce reduction, and the reputational effect of the owner's own public conduct toward advertisers.

Inflection 2 — the debt refinancing window (Jan–Apr 2025). The clearing of USD 11.2bn of hung debt in six weeks, at 97c to par rather than the deep discounts anticipated, was the single most important event in the company's post-acquisition financial history. It was driven less by improving fundamentals than by two exogenous factors: the perceived proximity of the owner to the incoming US administration, and the embedded optionality of X's then 10% stake in xAI, valued at approximately USD 5bn, which buyers of X paper received exposure to.

Inflection 3 — the mix inversion (2025–2026). Advertising stabilised at roughly USD 1.8bn while subscriptions added USD 365m in FY2025 and USD 177m in Q1 2026 alone. Twitter's FY2021 mix was approximately 89% advertising and 11% data licensing, with no subscription line. By FY2025, within the itemised segment sub-lines, advertising accounted for roughly a tenth and AI infrastructure plus data licensing for over half.

Inflection 4 — absorption (Mar 2025 and Feb 2026). Both mergers were all-stock, related-party transactions between entities under common control, executed without an arm's-length price discovery process. The March 2025 valuation of USD 33bn for X equity was 25% below the USD 44bn paid in 2022 and 25% above the USD 44bn internal round completed weeks earlier — an internal inconsistency the company has not explained.

22.1 Management guidance

No X-specific guidance exists. Parent guidance from the 10 August 2026 call: approximately USD 100bn annualised revenue run-rate by end-2026, and an internal USD 1tn revenue target advanced to 2030 from 2031. The S-1's X-specific qualitative commitments are to grow engagement, Premium conversion, advertising revenue per user, and advertiser diversity, and to complete the everything-app integration of information, communications, media, payments and banking.

22.2 Consensus expectations

There is no sell-side consensus on X. Consensus on SPCX is directionally positive with extreme dispersion (targets USD 200–600). Citi revised 2026 and 2027 forecasts upward after the Q2 beat. For X specifically, the only forward-looking anchors are the disclosed sequential subscription lift of USD 177m in Q1 2026 and the AI segment's 247% Q2 revenue growth — the latter almost entirely attributable to the Anthropic compute contract rather than to anything X does.

22.3 Bull case

1. The mix inversion is genuine and undervalued. Subscriptions added USD 365m in FY2025 and USD 177m in Q1 2026 alone. If that sequential run-rate holds, subscription revenue approaches parity with the USD 1.8bn advertising line within roughly three years, at structurally higher margin and with lower brand-safety beta. X would stop being an advertising business, which is precisely what its advertising problem requires.

2. X Money is a real option that costs almost nothing to hold. X enters consumer fintech with 4.4m existing paying subscribers, instant creator-payout integration, and a partner-bank structure requiring no charter. The 6% APY is a subsidised acquisition cost the parent can fund from USD 3.5bn of quarterly EBITDA. If even 10% of X Premium subscribers become active deposit customers, X acquires a fintech user base that would cost hundreds of millions to buy independently.

3. The data asset is systematically underpriced. Approximately 350m posts per day of fresh, timestamped, real-world human text is a scarce input in a market where AI labs are demonstrably willing to pay. Data licensing added USD 88m incrementally in FY2025 from what is almost certainly an under-commercialised base. Reddit has established the comparable. X's corpus is larger, faster and, for grounding purposes, more valuable.

22.4 Bear case

1. The regulatory cluster is correlated and potentially existential in Europe. Five proceedings across three jurisdictions opened within six weeks, substantially traceable to one product decision. The DSA permits fines to 6% of global turnover — and following the SpaceX merger, "global turnover" is a materially larger number than it was in December 2025. Ofcom holds business-disruption powers. X has chosen to litigate rather than settle, extending the exposure window and, if it loses, establishing precedent against itself.

2. The daily-usage trend invalidates the data thesis it is meant to support. Mobile DAU fell 7% YoY to 123.7m in July 2026 while Threads rose 21.3% to 147m. If X is a data asset first, its data flow is the asset — and the flow is contracting while the company simultaneously throttles free-tier posting by 98% to combat spam. Fewer posts from fewer daily users is a direct impairment of the only rationale that survived two mergers.

3. The forward revenue thesis is one counterparty on 90 days' notice. Approximately USD 15bn annualised from Anthropic against a USD 3.2bn FY2025 segment base, with a contractually firm minimum of approximately USD 3.75bn and a counterparty reported to be building its own capacity. Strip that contract out and the AI segment is a business generating roughly USD 3bn of revenue against a USD 6.4bn operating loss and USD 12.7bn of annual capex. The valuation at which X was folded into that structure was never independently tested.

22.5 Catalysts and monitorables, next 12 months

22.6 Analyst verdict (300 words)

The most important thing to understand about X Corp. in August 2026 is that it is no longer an investable asset, no longer a self-directing company, and — on the evidence of its own parent's prospectus — no longer primarily an advertising business. It is a consumer surface and a data faucet inside a rocket company that has decided its next trillion-dollar market is compute.

The S-1 settled the question that four years of leaks could not. X's advertising revenue was USD 1.8bn in FY2025: 39.9% of Twitter's FY2021 level, roughly 15% of the trajectory implied by the 2022 acquisition deck, and small enough that it now rounds to a rounding error inside a group generating USD 7.81bn a quarter. Paid conversion sits at 0.8%. The 2022 thesis — that Twitter was a monetisation failure a better owner could fix — has been tested for four years and has not been vindicated on any disclosed metric.

What has been vindicated is the strategic argument, and only narrowly. X's 350 million daily posts are a genuinely scarce input, and its distribution surface has successfully carried Grok to 117 million monthly users and X Money to national launch in under a year. Subscriptions are growing faster than advertising ever recovered. That is a real business, and it is a different one.

Against this sits a risk profile that is unusual in its correlation. Five regulatory proceedings across three jurisdictions, all traceable to one December 2025 product decision, with EU exposure now measured against a vastly larger turnover base. A CEO seat empty for thirteen months. Daily usage in structural decline behind Threads. And a forward revenue thesis resting on one customer with a ninety-day exit.

The verdict is not that X is failing. It is that X has been resolved into an input, and inputs do not get valued — they get consumed.

Compiled 14 August 2026. Every figure is tagged to a source tier; unverifiable data points are marked "not publicly disclosed" rather than estimated. Where sources conflict — notably on FY2025 advertising direction, Q2 2026 net loss, and the Yaccarino and Taylor appointment dates — both readings are stated with the discrepancy identified.

--