§ 0
The Foundation Principle
Every rule in this document is a downstream expression of one constraint:
The load-bearing rule
The sponsor's obligation is defined before the competition opens. The solver's obligation is defined at submission. Neither party may change their terms after the clock starts.
This is not a preference. It is the structural guarantee that makes the platform trustworthy to both sides. A sponsor who posts vague criteria and then cherry-picks is not a bad actor — they were allowed to post a bad challenge. The platform's job is to prevent that upstream, not adjudicate it downstream.
§ 2
Rules for Solvers
A Solver is any individual or team that submits a solution to an open challenge. There are no eligibility requirements beyond a verified account. Age, geography, credentials, employment status — none of these determine who can compete. The solution is judged, not the solver.
2.1 — Participation
Rule 2.1.a
Solvers may enter any open challenge. There is no entry fee. There is no minimum qualification. Pseudonymous participation is permitted through evaluation; verified identity is required only at payout.
Rule 2.1.b
Teams of any size may form and enter together. Teams self-organize; the platform does not assign teams. Team prize splits are the team's internal responsibility and must be declared at submission time.
Rule 2.1.c
Solvers may enter multiple challenges simultaneously. There is no restriction on parallel participation.
2.2 — Submission
Rule 2.2.a
At submission, solvers must provide: (1) the solution itself, (2) a hash commitment (see § 5), and (3) a declaration that the work is original and does not infringe on third-party IP. Knowingly false declarations result in permanent account suspension and forfeiture of any prize.
Rule 2.2.b
Solvers may use any tools, including AI systems, to produce their solutions. The solver is responsible for the output. "The AI made it" is not a defense against quality, originality, or IP claims.
Hard rule
Rule 2.2.c
Solvers may not submit solutions that contain third-party trade secrets, confidential information, or materials obtained through unauthorized access. Violations result in immediate disqualification and potential legal referral.
2.3 — Solver protections
Rule 2.3.a
Solvers retain all IP rights to non-winning submissions. The platform and the sponsor acquire no rights to any submission that does not receive a prize, regardless of how useful or relevant it may be to the sponsor.
Rule 2.3.b
Solver identities are never revealed to sponsors until after a winner is declared. Blind evaluation is enforced by the platform — not the honor system.
Rule 2.3.c
All submissions are cryptographically timestamped at receipt. This timestamp is the solver's proof of prior art, independent of the platform's continued existence. Solvers may download their timestamp receipt at any time.
§ 3
Intellectual Property Framework
IP terms set at launch determine participation rates. Strict acquisition terms suppress talent. The recommended default is Shared IP — and the data from two decades of prize competitions backs this.
| Template |
Solver keeps |
Sponsor gets |
| Shared IP (Default) |
Full ownership of their solution |
Royalty-free, non-exclusive license to use the winning solution |
| Solver Retains |
Full ownership; no license granted |
Right to evaluate; no implementation rights without separate agreement |
| Exclusive License |
Ownership; exclusive license granted for defined period and scope |
Time-limited exclusivity (e.g., 3 years, healthcare sector only) |
| Sponsor Acquires |
Nothing — full transfer on prize payment |
Full ownership, all rights |
| Open Source |
Public recognition; no commercial exclusivity |
Public use; sponsor named as funder |
Sponsor advisory: The instinct to acquire IP is understandable but usually counterproductive. What most sponsors actually need is the right to implement — not exclusivity of ownership. Shared IP grants that right while keeping the solver ecosystem healthy. If your prize is $5,000, you are not buying a patent. You are buying a solved problem.
§ 4
Defining "Solved"
The most contested word in any challenge platform. The answer here is structural: the sponsor defines it, the platform enforces the definition, and neither party can move the goalposts after the challenge opens.
Type A — Quantitative
Problems with measurable, objectively scorable outputs. Examples: reduce error rate below 3%, process 10,000 records in under 2 seconds, achieve prediction accuracy above 92% on the test set.
Platform scores submissions against defined benchmarks automatically. No human judgment, no subjectivity, no delay. A submission either clears the threshold or it doesn't — no award for one that fails, no denial for one that clears.
Type B — Deliverable
Problems requiring a defined artifact: a working API with these endpoints, a 10-slide deck addressing these four questions, a 500-word policy memo structured around these requirements.
Sponsor defines a public checklist at launch. Evaluation is whether the submission satisfies each item. Sponsors may not add checklist items after launch. If a sponsor claims a deliverable fails, they must identify which checklist item it fails and provide evidence.
Type C — Qualitative-Judged
Problems requiring judgment: best strategic recommendation, most compelling creative concept, strongest risk analysis.
Sponsor publishes a weighted scoring rubric at launch (e.g., feasibility 40%, clarity 30%, originality 30%). For prizes above a defined threshold, a third-party review panel applies the rubric. Solvers may challenge the scoring application — not the rubric itself.
4.2 — The no-award option
Sponsors retain the right to award no prize if no submission meets the defined success criteria.
Hard rule
Rule 4.2.a
No-award requires written justification mapped to the original criteria. Sponsors cannot invoke no-award because they "weren't impressed" or because a better solution exists internally. They can invoke it only if submissions failed to meet the posted criteria. The justification is published publicly. Solvers may dispute it.
Repeated no-award invocations by a sponsor degrade their reputation score and eventually require a third-party panel for all their challenges. This is market enforcement — not a penalty, but a trust signal the platform surfaces.
§ 5
Submission Integrity & Cryptographic Proof
The "sponsor steals the work" problem is the trust-killer for any challenge platform. The solution is technical, not contractual. We use a cryptographic commitment scheme — the same approach used in secure auctions and zero-knowledge proofs.
01
Commit
When a solver submits before the deadline, the platform generates a hash of their solution combined with a platform-generated salt: commit = SHA-256(solution + salt). The commit hash is recorded with a tamper-evident timestamp and given to the solver as a receipt. The solution itself is encrypted at rest — the sponsor cannot see it.
02
Close
When the challenge deadline passes, submissions are locked. The commit hashes for all submissions are published in a public log — independent of the platform, verifiable by anyone.
03
Reveal
Encrypted solutions are decrypted and made available to the sponsor for evaluation. The platform verifies that each decrypted solution matches its previously published commit hash — proving the solution was not modified after the deadline.
04
Solver receipt
Each solver receives a downloadable proof bundle: their submission hash, the platform's timestamp, and a signed certificate. This is their independent prior-art record, usable outside the platform in any legal proceeding.
Platform transparency commitment: The commit log for every closed challenge is permanently public. If the platform ceases to operate, the commit logs remain accessible. Solvers' proof of prior work does not depend on the platform's survival.
§ 6
Dispute Resolution
Disputes arise in one of three forms: (1) a sponsor claims a solution doesn't meet criteria; (2) a solver claims their winning solution was not awarded; (3) a solver claims their non-winning solution was used without compensation. Each follows a defined track.
Track A — Sponsor disputes a solution
Sponsor claims the submitted solution does not meet the posted success criteria.
01
Sponsor files specific objection
Sponsor submits a written objection identifying: (a) the specific criterion the submission fails, (b) why it fails, citing evidence from the submission. General objections ("not what we needed") are not accepted.
Day 0–3 of Review Period
02
Solver response window
The solver receives the objection (with sponsor identity redacted if they choose) and has 48 hours to respond: rebut with evidence, acknowledge the failure and withdraw, or request escalation to arbitration.
Day 3–5
03
Platform mediation
A platform mediator reviews the objection, the criteria, and the response. For Type A and B challenges, the mediator applies the criteria mechanically. For Type C, the mediator assesses whether the rubric was applied as written. A preliminary finding is issued.
Day 5–7
04
Third-party arbitration or outcome executed
If either party rejects the preliminary finding: an independent arbitrator with domain expertise reviews the full record. Decision is final and binding. Arbitration cost is split equally unless a party acted in bad faith. If preliminary finding is accepted: prize is released or no-award is confirmed.
If escalated
Solver prevails
Prize released from escrow in full. Sponsor reputation score decremented.
Sponsor prevails
No prize awarded. Escrow returned minus hosting fee. Solver retains all IP in their submission.
Track B — Solver claims wrongful non-award
Solver believes their submission met the posted criteria but was not awarded a prize.
Solver files a criteria-mapped appeal within 5 days of the sponsor's evaluation. The appeal must identify which criteria the solution satisfies and provide evidence. If the platform mediator finds the solution meets criteria, the burden of proof shifts to the sponsor to demonstrate non-compliance.
Note on qualitative challenges: Track B disputes on Type C challenges are limited to procedural review — was the rubric applied? Were all submissions reviewed? Solvers cannot appeal the substance of a judgment that was correctly applied.
Track C — Unauthorized use of a non-winning submission
A solver believes their non-winning submission was used by the sponsor without authorization.
Solver files a claim with evidence of unauthorized use. The platform provides the solver's cryptographic timestamp receipt as independent proof of prior creation. The platform refers the dispute to legal arbitration — this is beyond administrative resolution.
The platform's role in Track C disputes is limited to: (1) confirming timestamp records, (2) revoking the sponsor's platform access pending investigation, and (3) notifying the solver of their right to pursue external legal remedies. The platform is not a court and does not adjudicate IP claims.
§ 8
What Cannot Be Posted
The platform does not curate by industry or topic. Any real problem with a real prize can be listed. The following are structural exclusions — not preferences.
8.1 — Verifiably harmful
Challenges whose solutions would primarily or foreseeably enable mass harm: weapons systems, surveillance infrastructure, content designed to deceive at scale, or anything requiring solvers to produce material that would be illegal to possess or distribute.
8.2 — Non-problems
Challenges that are not real business problems: market research disguised as challenges, opinion gathering with no prize intent, challenges designed to generate free labor for tasks the sponsor intends to do regardless of outcome.
8.3 — Unverifiable success
Challenges where the success criteria are inherently unknowable at posting time (e.g., "increase our revenue over the next year") cannot be listed. Success must be determinable within the competition window.
Note on sensitive industries: Healthcare, legal, and financial challenges are permitted but must include a disclaimer that winning solutions do not constitute professional advice and may require regulatory review before implementation.
Sources & references
HeroX IP Guide — five-template IP model; 90% of successful challenges use Shared IP or Innovator Retains model; correlation between strict IP terms and reduced participation.
InnoCentive / Wazoku — Challenge Specific Agreement as binding pre-competition contract; non-exclusive licensing as default; $60M+ awarded across thousands of solvers as proof-of-model.
15 U.S.C. § 3719 (America COMPETES Act) — federal principle that government may not gain IP interest without written consent; blind judging requirements; judge independence provisions.
Cryptographic commitment literature — SHA-256 hash-based commit/reveal protocol; hiding and binding properties. Ferreira & Weinberg (2020), Princeton.
challenge.gov / GSA — criteria-first challenge design; distinction between prize competitions and procurement contracts.
Wolflow AI LLC — this document. Concept originated May 2026. Not a legal document. Consult counsel before implementing any provision in a live platform.