# Entity.ID — Investor FAQ / Objection Bank 25 hardest questions, strongest honest answers. Facts carry inline sources (full URLs in `brand/research/`). `[FILL: …]` = only Husam can answer — insert before use. Answers are written to be spoken; trim freely. **Positioning line to return to whenever cornered:** *Entity.ID is compliance infrastructure that exists before the mandate — when anyone asks "who answers for this agent?" or "who owns this venture?", it's the system already holding the answer.* --- ## Category & product ### 1. "Isn't this just Stripe Atlas?" No — Atlas ends where we begin. Atlas is a Delaware filing plus onboarding into Stripe's payments stack: nothing exists before the state filing, it's US/Delaware-only, and KYC is bound to a natural person — humans only. Entity.ID starts *before* the filing: a venture is named, membered, owned, and operating (treasury, governance, fundraising) in minutes, with incorporation as a later option. And agents can be registered actors, which Atlas structurally cannot do. Atlas is actually our favorite evidence: its incorporations grew **+130% YoY in Q1 2026** (Stripe/MindStudio) — AI is minting founders faster than the paperwork system was designed for. That demand spills straight into "operate first, incorporate when ready." ### 2. "What do you actually do, in one sentence?" Anyone — founders, collectives, and AI agents — forms a real venture in minutes: a named entity with members, ownership, rules, and a treasury, recorded in a tamper-proof public registry, operating immediately, incorporating later if needed. ### 3. "What's real today vs. roadmap?" Real, shippable, shown in the demo: the 7-step formation flow (name → details → partners → ownership split → arbitration terms → constitution → register), the public venture record and address (`name.public.entity.id`), and the dashboard modules — overview, treasury, fundraising, governance, members, compliance. Roadmap: [FILL: Husam to confirm the honest split — e.g. incorporation bridge, agent KYA API surface, arbitration integration depth, which treasury rails are live]. The demo venture (Tarka Labs) is populated fictional data and we always label it that way — we don't dress the demo as traction. ### 4. "Is the demo real?" The screens are the real product's screens; the data is a fictional demo venture, clearly labeled. What's live for the public today is app.entity.id. [FILL: one sentence on exactly what a stranger can do on app.entity.id today.] --- ## Legal reality ### 5. "Is any of this legally real?" Yes — deliberately assembled only from parts that already work. Machine-formed contracts have been binding in the US since 1999 (UETA §14, adopted in 49 states, plus the federal E-SIGN Act 2000). US LLC law can already put an algorithm in operational control of an entity — a decade of scholarship from Shawn Bayern (Northwestern L. Rev. 2014 onward). Wyoming charters algorithmically managed entities by statute (DAO LLC 2021; DUNA effective July 2024, now also Alabama and West Virginia, April 2026). And in May 2026 an autonomous agent reportedly formed its own LLC, got an EIN, and opened an FDIC-insured bank account (techstartups.com — say "reported"). We sell only what's legal today: the entity wrapper plus a disclosed human accountability chain. Direct AI personhood — rights for the machine itself — is politically dead, and we don't need it. ### 6. "Are these ventures recognized by courts and banks before incorporation?" Honest answer: a pre-incorporation venture on Entity.ID is a structured agreement among its members — and agreements bind. The record gives it what informal teams never have: named parties with verified identity, an agreed ownership split, working rules, and an arbitration clause chosen *before* any dispute exists. That's contract-law enforceability plus evidentiary quality, not state personhood — and when the venture needs state recognition, the record maps onto a traditional filing ("incorporate when you need to"). Compare the default: teams operating on handshakes are involuntary general partnerships — a US court held DAO participants personally liable as general partners in *Samuels v. Lido DAO* (2025). Unstructured is not a neutral choice; it's the worst structure. ### 7. "Who's liable when a registered agent causes harm?" The chain on the record: agent → venture → members/owners → accountable humans or companies. That's the point. California's AB 316 (effective Jan 1, 2026) already says "the AI did it autonomously" is not a defense — deployers own their agents' conduct by statute; what they lack is the instrument that structures and discloses that ownership. Recourse is built in: arbitration ships in the formation terms, and a disclosed, structured entity is an insurable object — agent-liability insurers (Armilla, Testudo — both live 2026) currently have no standardized thing to underwrite. ### 8. "Why will regulators accept a private registry?" We don't ask them to defer to us — we hand them the supervision surface they currently lack. The precedent is exact: after 2008, the G20 created the LEI because regulators couldn't identify trade counterparties; 3M+ entities now carry one and it's compulsory in derivatives reporting (GLEIF). The demand signals are already in print: the FSB's June 2026 consultation treats agents as "synthetic employees" needing clear accountability chains; FINRA's 2026 Regulatory Oversight Report flags agent autonomy and traceability; NIST ran an RFI on AI-agent identity (March 2026); EU AI Act Article 50 disclosure becomes enforceable **August 2, 2026**, with Commission draft guidance explicitly extending it to agents. Regulation fills gaps with whatever infrastructure exists when the rule is written. We're building the thing that's standing there on that day. ### 9. "Isn't AI personhood scary / politically toxic?" It would be, if we were doing it. We grant nothing to the machine. The winning route — and the only one law allows — is *entity* law: the agent registers as or within a venture anchored to disclosed humans or companies, becoming a subject of obligations, not a rights-holder. The EU proposed "electronic persons" in 2017 and retreated; several US states are now banning AI personhood outright (NPR, May 2026). Those bans are our moat, not our threat: they foreclose every route to agent standing *except* entity-with-human-anchors — which no personhood ban touches. And note the dark-mirror argument: the scariest object in the legal literature is LoPucki's "algorithmic entity" with no traceable human controllers. That fear is precisely the demand curve for a disclosure-first registry. --- ## Competition ### 10. "What stops Stripe or Carta from doing this?" Three structural things. **Neutrality:** a registry owned by a payments company is that company's ledger — Visa won't gate on Mastercard's attestations, platforms won't trust a competitor's registry; the standing layer structurally wants a neutral open standard, which is what the Entity.ID protocol is. **Incumbent economics:** Stripe's and Carta's revenue starts *after* incorporation (Carta Launch literally requires a certificate of incorporation to onboard); starting before the filing reframes their own funnel. **Legal texture:** formation semantics — ownership splits, constitutions, arbitration clauses, disclosure chains — are years of domain depth a payments team won't rebuild. Also worth saying: Carta's January 2024 cap-table data scandal (an employee pitched a customer's investors using confidential data; Carta exited secondaries "to prioritize trust") is the standing argument for a public record over a vendor's private database. ### 11. "MIDAO and Otoco already market 'AI agent incorporation.'" They validate the category and prove the gap. MIDAO charges **$9,500 + $2,000–5,000/yr**, onboards over Telegram in weeks, and mandates human oversight with kill switches — the agent is never the member or principal (midao.org). Otoco lets an agent mint a $99–299/yr wrapper — the agent as customer of paperwork, not as registered actor. Nobody puts agents *inside* entities as members with treasury mandates and governance-scoped permissions. That's the whole product. ### 12. "How are you different from Skyfire / Catena / the KYA startups?" They verify the *operator*; we register the *actor*. Skyfire's KYA vets an agent's developer and issues a payment credential. Catena Labs ($30M Series A, May 2026, applying for an OCC trust charter) banks agents under Catena's own charter — the standing belongs to Catena, never to the agent. Every KYA-adjacent player chains the agent back to a *pre-existing human-owned entity*; nobody offers the inverse — entity, ownership, treasury, and governance native to the agent. Catena is a natural partner, not a competitor: an agent-native bank wants registered ventures as account holders. ### 13. "What about the payment networks' agent protocols?" Complementary, and they prove our layer is missing. Each rail answers one narrow question: Google AP2 mandates prove *a human authorized this purchase*; Cloudflare's signed agents prove *which operator sent this request*; Visa TAP and Mastercard Agentic Tokens scope *spend*; Microsoft Entra gives agents *logins inside one tenant*. None makes the agent a *party* — with owners, a treasury, rules, and a jurisdiction. All four payment stacks shipped agent rails within 18 months (Apr 2025–Jun 2026), and each presupposes a legally accountable party behind the agent that none of them provides. We're the presupposed layer. ### 14. "Your real competitor is 'do nothing' — why won't people just stay informal or grab an LLC?" Do-nothing is our biggest segment, not our competitor. 30.4M US businesses have zero employees ($1.8T receipts, Census 2025 release) and mostly never incorporate — the LLC path costs money, admin, and franchise tax for structure they use one page of. But informal isn't free: it's an involuntary general partnership with unlimited personal liability (*Samuels v. Lido DAO*, 2025), unwritten ownership, and no treasury. We price the first real structure at approximately the effort of a group chat. [FILL: actual pricing/free tier.] And for the second audience — agents — there is no do-nothing option at all: no legal existence is the current state. ### 15. "Why won't a government just build this?" Governments *sell* registries — that's our evidence, not our risk. Estonia's e-Residency: 132,000+ e-residents, 5,556 companies formed in 2025, **€125M state revenue in 2025** (e-resident.gov.ee / ERR). Delaware runs 25–30% of its General Fund on being a good registry. But each is single-jurisdiction by construction, moves at legislative speed, and none admits agents. The internet-native layer that interoperates *above* the state patchwork — and hands each jurisdiction its rules as machine-readable properties — is not a thing a single state can build. The EU's own AI Office concedes its agent thinking is "only preliminary" (2026). --- ## Market ### 16. "How big is this really?" Two engines, sized separately. **Ventures:** record ~5.62M US business applications in 2025 and accelerating (Census BFS); 30.4M US nonemployer businesses; 72.9M independent workers (MBO Partners 2025). The paperwork-only proxy: LegalZoom makes $756M/yr filing forms in a ~4%-CAGR market; the operations layer is where revenue concentrates (Carta ~$442M ARR). **Agents:** Gartner (Nov 2025) — by 2028, 90% of B2B buying mediated by agents, over **$15T of B2B spend through agent exchanges**; ~69k agents already transact ~$600M/yr over one payment protocol alone (Coinbase x402, spring 2026). We're the entity layer both engines are missing. Honest floor: even eMarketer's conservative agentic-commerce case is $144B by 2030 — the spread between $144B and $15T is itself the story: everyone agrees on direction, nobody can size a market that has no trust layer yet. ### 17. "Agentic AI might be hype. What if it stalls?" Then we're a venture-formation and operations company in a market setting formation records every year — the proto-entity business stands on its own (see #16). We also quote Gartner's bear case ourselves: >40% of agentic-AI projects canceled by end-2027 amid "agent washing." A shakeout is *good* for us — what survives a trust crisis is precisely the verifiable, accountable agent, and trust crises are historically what create registries (the LEI was born from one). ### 18. "Isn't the registry a cold-start problem — a registry of two?" Registries start as member clubs — SWIFT, Visa, and the LEI all did. Our wedge is that the formation product is single-player-useful: a venture gets treasury, governance, and record value with zero network. Those ventures then become the accountable anchors the agent side needs — the first verification corridor is agents transacting *inside and between* Entity.ID ventures. Every registered entity makes the next standing query more valuable; when the queriers are also machines, the D&B flywheel that took decades compounds in quarters. ### 19. "What's the moat?" Four compounding ones. (1) **Registries compound** — Delaware/D-U-N-S/LEI dynamics at API speed. (2) **Neutrality** — platform IAM stops at the tenant edge and payment networks won't trust each other; a neutral open standard is structurally required, and open standards are winner-take-most. (3) **Legal texture** — years of formation-semantics depth. (4) **Two-sided lock-in** — the human ventures the formation product wins are the accountability anchors the agent product needs; both live on one registry. --- ## Business ### 20. "How do you make money?" [FILL: Husam — pick the primary model and price points. Grounded candidates, with comps from the research:] - Registration + renewal (the Delaware/Estonia model — one state earns ~$1.8–1.9B/yr in franchise fees; Estonia's registry earned €125M in 2025). - Verification/standing queries as an API (the D-U-N-S model — the file counterparties pay to read). - Operations take: treasury, fundraising, and incorporation-upgrade fees (Carta's ~$442M ARR shows operations is where formation revenue actually lives). - KYA verification subscriptions for platforms and enterprises that must check agents at scale. ### 21. "What's the traction?" [FILL: Husam — real numbers only: registered ventures, app.entity.id signups, waitlist, LOIs, pilot conversations. If pre-traction, say so and pivot to demo + why-now: the product is built and the mandate (EU Art. 50) has a date next month.] ### 22. "Why you / who's the team?" [FILL: Husam — team slide facts: founders, relevant background (registry/legal/fintech/AI), advisors, and the story of why this team saw the entity gap first.] ### 23. "What are you raising, and what does it buy?" [FILL: Husam — ask, instrument, valuation/cap, runway, and 3 milestones it buys (suggested shape: (1) live incorporation bridge, (2) first agent-KYA API partner, (3) N registered ventures by date).] --- ## The awkward ones ### 24. "What's the underlying tech?" — say-this / not-that table This one gets asked with 2022-flavored vocabulary. Per the brand's terminology rules, answer in these words and never the ones on the right: | Say this | Not that | |---|---| | a tamper-proof public registry | any 2022-era infrastructure brand names | | tamper-proof records; verifiable | "on-chain" | | programmable rules | "smart contracts" | | the open registry standard / the Entity.ID protocol | the retired project name | | "independent, neutral arbitration service (Kleros v2)" | naming the arbitration layer's own tech stack | If pressed on architecture: "The registry is an open standard; records are publicly verifiable and can't be silently edited — the record, not the vendor, is the trust anchor. Happy to go deeper in diligence." Then actually go deeper in diligence — under NDA, with the engineering story. [FILL: Husam — decide the diligence-depth answer on infrastructure.] ### 25. "This sounds like a legal opinion with a UI. What's defensible about software here?" The registry *is* the software: identity verification at membership, live treasuries with signing policies, weighted governance with quorum, fundraising instruments settling to treasury, machine-readable standing checks — that's an operating system, not a document generator. The document generators (LegalZoom, Clerky) are exactly what we're not: they stop at the filing; we run the venture. And the agent surface — a standing check a counterparty machine can run before value moves — has no paper equivalent at all. --- ## Rapid-fire one-liners (memorize five) - **Agents can pay. They can't be anyone.** (V4) - **Incorporation is a feature, not the front door.** (V6) - Every payment network shipped agent rails in 18 months — and each one presupposes a legally accountable party that none of them provides. - An airline argued in court that its chatbot was "a separate legal entity." It lost — because no such entity existed. We make the honest version of that sentence true. (*Moffatt v. Air Canada*, 2024) - When regulators ask "who answers for this agent?" — we're the system already holding the answer.