Reviewer Agreement

1. Introduction and acceptance## 1. Introduction and acceptance

1.1 · The parties. This Reviewer Agreement (the "Agreement") is a binding contract between XSci LTD, a company registered in England and Wales (66 Paul Street, London, EC2A 4NA, United Kingdom) that operates the XSci platform ("XSci", "we", "us", "our"), and the individual who accepts it (the "Reviewer", "you", "your").

1.2 · When you accept. You accept this Agreement when you apply to join the XSci reviewer program and, thereafter, each time you accept a review assignment. If you do not agree, you may not act as a reviewer for XSci.

1.3 · What it governs, and what it is read with. This Agreement governs your relationship with us as a reviewer. It supplements — and should be read together with — our Certification Standard, Certification Procedure, Peer Review Policy, AI and Data Usage Disclosure, Privacy Policy, and Terms of Service.

1.4 · Which document controls on a conflict. Where this Agreement conflicts with those documents on the terms of your engagement — your status as a contractor, compensation, confidentiality, intellectual property in your report, term and termination — this Agreement controls. Where the conflict concerns what an assessment decides, the criteria applied to it, the form a report must take, or your conduct in reaching a decision, the Certification Standard and the Peer Review Policy control, and this Agreement is to be read as giving effect to them.

2. Your role and standards

2.1 · What you are asked to decide. You conduct each assessment as a qualified, independent expert, applying the criteria set out in the Certification Standard. Certification requires that both of 2.2 and 2.3 hold, and they are independent of one another.

2.2 · (a) No major error. A major error is a defect which, while it stands, means the manuscript's central claims are wrong, unsupported, or impossible to evaluate. A defect that ordinary revision would put right is not a major error.

2.3 · (b) A genuine prospect of publication, after ordinary revision, in the venue the authors are aiming at. Up to three target journals may be named at purchase, and the test is met if the manuscript would fit any one of them. Where no target is named, the target is the leading journals of the manuscript's field, and you must identify and name them in your report.

2.4 · The bar is relative to the target named, not absolute. A sound manuscript aimed higher than it can reach is not certified, and you must recommend where the work would fit instead.

2.5 · The form of your decision. You record exactly one of two outcomes: certify, or do not certify, because — with every reason stated specifically. There is no score, grade, rating, band or ranking, and "revise and I will judge then" is not an available outcome. For each of the nine assessment domains the manuscript engages you record exactly one of four findings — no concern, concern noted but not material, material concern, or not applicable with reason — with at least one sentence specific to that manuscript. You must not sum, average, or otherwise convert those findings into a number, letter or fraction. Not applicable describes the manuscript, not you: record it only where the manuscript does not engage the domain at all, and it is never a mark against the work. Where you cannot assess a domain the manuscript does engage, you must tell us rather than record it as not applicable.

2.6 · Fit is a required section of every report. Your judgement on fit, and the recommendation that follows from it, are kept separate from the domain findings.

2.7 · Limits on what you may ask of an author. You must not:

(a) condition certification on any further purchase, or on the author naming a different target venue, or offer to reconsider if they do — your recommendation on fit is advice and never a condition; (b) contact an author directly, or accept contact from one. All communication passes through the report and the author response mechanism; (c) assume credit in the submission for the assessment. Assessing a manuscript is paid work and is never authorship, contributorship, or acknowledgement.

2.8 · Our obligations to you in return. No member of XSci staff may ask you to change a decision, and none has the authority to do so. Your report is checked by us for completeness only, never for its conclusion (clause 8).

2.9 · Other standards. You are engaged as an independent contractor, not an employee, agent, or partner of XSci. Nothing in this Agreement creates an employment or joint-venture relationship. You are responsible for your own taxes and statutory contributions. You accept assignments only where you hold genuine subject-matter expertise, and complete them within the timelines communicated to you — ordinarily five to ten days from assignment, per the Peer Review Policy. You will follow COPE ethical guidelines for peer reviewers and conduct yourself professionally, constructively, and without discriminatory or abusive language.

3. Named accountability and use of AI by the reviewer

3.1 · What the service is. XSci offers Human (Expert) Review as a service in which a named human expert takes full responsibility for the report delivered to the author, and in which no machine certifies anything. Clauses 3.2 to 3.6 follow from that.

3.2 · You are fully responsible for the content of every report you submit, including its conclusions and recommendations, as the accountable human reviewer. The certification decision is yours and cannot be produced by any automated system, alone or in combination.

3.3 · Authorised AI tools. You may use AI tools authorised by XSci to help draft or structure your report. If you do, you must personally verify every substantive statement and, before submitting, confirm (via the required checkbox) that you take final responsibility for the report.

3.4 · What you must not do. You must not submit AI-generated output as your own independent expert judgment without meaningful review, and you must not represent to authors, or allow XSci to represent, that a Human Review was produced "without AI" or by "purely manual" means. The commitment we make to authors is accountability ("a named person is responsible"), not the absence of AI.

3.5 · No external tools. You must not paste manuscripts or any confidential material into external or personal AI tools, browser extensions, or third-party services (see clause 4).

3.6 · Where an AI review report exists for a manuscript, you choose whether to see it. Your choice is recorded on the assessment record, as is your own use of AI in preparing your report. An AI review report is not a verification — it is a generated narrative report, not a mechanical test — and it does not bind your assessment in any respect.

4. Confidentiality

4.1 · What is and is not confidential. Human review on XSci is purchased only for manuscripts already deposited publicly on the platform, so the manuscript itself is ordinarily public and this Agreement does not ask you to treat public material as secret. What is entrusted to you in confidence is the assessment, and you agree to 4.2 to 4.7.

4.2 · You will treat as strictly confidential: your report before it is released; the existence and content of the assessment; the author's response; any material the author supplies to you that does not form part of the public deposit, including underlying materials provided privately; and any communication with us about the assignment.

4.3 · You will treat as unpublished, confidential information any manuscript or version that is not publicly deposited, applying the standard for unpublished work.

4.4 · You will not disclose, share, forward, or discuss the assessment or any confidential material with any third party, and will not delegate the assessment to another person, without our prior written consent.

4.5 · You will not use any non-public information obtained through the assessment for your own research, commercial advantage, or any purpose other than performing the assessment.

4.6 · You will not transmit any confidential material to a third-party service (including external AI tools) that is not authorized by XSci under our supplier confidentiality and data-protection arrangements.

4.7 · You will securely delete local copies of confidential materials after your report is submitted, retaining nothing beyond what XSci stores on the platform.

4.8 · Your identity. Reports are delivered under XSci's default single-blind arrangement: your identity is not disclosed to the author, and we do not disclose it in a published report. Nothing is ever published about you individually, including in the aggregate figures XSci publishes under the Peer Review Policy.

4.9 · Survival. These confidentiality obligations survive the completion of any assessment and the termination of this Agreement.

5. Conflicts of interest

5.1 Interests that bar you from an assignment

5.1.1 · You must decline an assignment where any of the following applies:

(a) a prior role in this work — you have advised on the manuscript, reviewed it for a journal, or taken part in producing or correcting any part of it. There is one exception: if you assessed a version that was certified, you may assess a later version of the same manuscript; (b) a close personal relationship, or a dispute, with any author, of a kind a reasonable person would consider material; (c) a financial or commercial interest in the outcome, or in the subject matter to the extent that the outcome would affect it; (d) a current supervisory relationship in either direction with any author.

5.2 Interests you declare, which do not bar you

5.2.1 · Everything else is declared and is not a bar. This includes, and is not limited to: recent co-authorship or collaboration; a shared current institutional affiliation; competing or directly overlapping research interests; academic rivalry; and strongly held prior public positions on the question at issue.

5.2.2 · How you declare. You declare these to us before accepting. The declaration goes on the assessment record, and we decide whether to proceed. Declaring an interest under this clause is not a reason to decline and does not count against you.

5.3 An interest discovered after you accept

5.3.1 · A barred interest. If an interest under clause 5.1 emerges after you accept an assignment, you must stop work and disclose it to us immediately. The assignment is reassigned. You bear no consequence for doing so — it does not affect your standing in the reviewer program, your eligibility for future assignments, or payment for work properly done before you stopped.

5.3.2 · A declarable interest. If an interest under clause 5.2 emerges, notify us promptly and we will decide whether to proceed.

5.3.3 · What happens if a barred interest is never disclosed. Where an assessment has been completed and it later emerges that you held an interest that barred you under clause 5.1 and did not declare it, the consequence falls on the record, not on your report's contents: the certification is revoked under the Peer Review Policy, because a record that does not disclose such an interest is materially false, and the author is refunded. No replacement reviewer is appointed and the version is not assessed again — the refund is the whole of the remedy. Nothing about you individually is published, then or at any time. Separately, an undisclosed barred interest is a breach of this Agreement and may end your engagement under Section 12.

5.3.4 · Disclosing an interest — at any stage, including late — is always treated better than not disclosing it.

6. Review reports — licence, delivery, and intellectual property

6.1 · Why a licence is needed. XSci commissions and pays for expert assessments and delivers them to authors as part of its paid service. Clauses 6.2 to 6.6 make that possible.

6.2 · The licence you grant. You grant XSci LTD a perpetual, worldwide, non-exclusive, royalty-free, sublicensable licence to store, reproduce, edit for format, display, distribute, and deliver your report (and any refinements or revised versions) to the relevant author(s), to publish it where the author has elected disclosure under the Peer Review Policy, and to retain it as part of the manuscript's assessment record.

6.3 · Your report may be published. The purchaser elects at purchase whether the report will be disclosed, and a report bought at the discounted price is published automatically 48 hours after it is released to the author. The author may also open a privately bought report at any later date, and that right is not limited in time — a report bought privately today may be published by the author years from now, without notice to you and without your agreement being sought. You will be told, at assignment, whether disclosure has been elected, but the author may change that election afterwards and you cannot make disclosure a condition of accepting an assignment. Write every report on the basis that it may become public, under the single-blind arrangement in clause 4.8.

6.4 · You retain authorship of your own report. This Agreement does not transfer copyright in your underlying expertise; it grants the licence necessary for XSci to operate the service, deliver your report, and publish it where the author has so elected.

6.5 · A judgement in a delivered report is never altered. Where you wish to add or correct something after delivery, you do so as a dated addition, and the original is retained alongside it.

6.6 · No separate publication. You will not separately publish, sell, or redistribute the report you produced for XSci, as it was commissioned for delivery through the platform.

7. Improving XSci's review systems

7.1 · The use, and why it is not optional. To improve the quality, consistency, and speed of review, XSci uses the reports you submit and your refinements/edits to AI-assisted drafts (i.e., the paired data of AI output and your expert corrections) to evaluate, train, and improve XSci's own review systems. This use is part of the paid engagement and is not separately optional, because it is the only means by which XSci can tell whether the review systems its authors rely on are working. You are not named — see 7.3.

7.2 · Scope — reviewer outputs only. This covers your review outputs and edits only. It does not extend to an author's confidential material: consistent with our commitment to authors, the manuscript, and anything else an author supplies in order to obtain a service and does not publish, is never used to train, fine-tune or evaluate any model. Published content is governed by the licence authors grant in the Terms of Service and by the AI and Data Usage Disclosure.

7.3 · You are not named. Nothing XSci publishes about review quality identifies you — not aggregate findings, not released datasets, not case material, and not any public description of this work. A reviewer's right to remain anonymous is a longstanding principle of peer review and XSci keeps it intact.

7.4 · Anonymity is about attribution, not about use. Your reports are still used under 7.1; internal evaluation and improvement proceed on an anonymised basis.

7.5 · Your anonymity never affects your standing in the reviewer program, your eligibility for assignments, or your compensation.

7.6 · Governed by. This use is governed by our AI and Data Usage Disclosure ("How We Improve Our Review Systems") and the DPA, and is limited to XSci's own systems and infrastructure.

8. Compensation

8.1 · The reviewer payment. XSci pays reviewers a reviewer payment for completed assessments, at the rates and on the terms communicated to you at assignment or in a separate payment schedule. Two amounts are kept apart throughout our documents and should not be confused: the assessment fee is what an author pays XSci for an assessment, and the reviewer payment is what XSci pays you for carrying it out. Neither is calculated from the other, and what an author paid is not a measure of what you are owed.

8.2 · Your payment does not depend on your decision. No part of your compensation varies according to whether you certified, and a refusal is a completed, payable assessment, not a failure of the service. Your payment does not vary with the assessment fee the author paid, with whether the report is disclosed, or with the round number.

8.3 · Payment turns on how the assessment was carried out, never on what you decided. Fees are payable for assessments completed and submitted in accordance with this Agreement and the Peer Review Policy. A report that is incomplete, plagiarised, or otherwise non-compliant with those documents is not a completed assessment and is not eligible for payment, because the author has not received what they bought.

8.4 · Where an appeal by the author succeeds, the reviewer payment is not due. An appeal lies on four grounds only — the Certification Procedure was not followed; you held an interest that barred you; the report contains an error of fact about the manuscript; the report falls below the required standard of detail — and each of them is a finding that the assessment was not delivered as required. Where an appeal succeeds on any of them, the payment for that assessment is not payable and, if it has already been paid, it is recoverable — by set-off against sums due to you on other assignments, or as a debt.

8.5 · You are told the ground and you may answer it before anything is withheld or recovered. Where a ground is raised that could affect your payment, we put it to you, identify what is said to be wrong, and give you a proper opportunity to respond. There is no limit on the number of exchanges. A question of fact in particular is often resolved this way rather than decided: where it is resolved, the appeal does not succeed and your payment stands. Where it cannot be resolved between you and the author, XSci adjudicates, and may take the view of an independent expert at its own cost before doing so.

8.6 · A decision we disagree with is never a ground for withholding payment, and neither is a refusal to certify: a refusal delivered in full is a completed, payable assessment. What is at issue is only ever whether the assessment was carried out and reported as this Agreement requires. Your report stands on the record as you wrote it: an appeal does not rewrite it, withdraw it, or change your decision.

8.7 · The completeness check. Before a report reaches the author we check it for completeness against the Certification Procedure — never for its conclusion. An incomplete report is returned to you with the omission identified, up to three times; after that the assignment is reassigned. Time spent on a returned report is part of the assignment and is not separately charged or deducted.

8.8 · How payment is made. Through XSci's payout provider (Stripe Connect); you are responsible for completing onboarding and for any taxes on amounts received.

8.9 · Where an assessment is closed because you referred an integrity matter under clause 5.3 or the Certification Procedure and stopped, you are paid for work properly done before you stopped, and nothing is recovered from you. The author is refunded because no report was delivered, and that refund is borne by XSci: referring a matter you are required to refer never costs you money.

8.10 · Where an assessment is reassigned because you stopped on discovering a barred interest under clause 5.3, you are paid for work properly done before you stopped.

9. Representations and warranties

9.1 · You represent and warrant that:

(a) the assessment is your own work, reflecting your genuine expert judgment (with any AI assistance reviewed and owned by you per clause 3); (b) you have the qualifications and expertise you have represented to XSci; (c) you have declined where clause 5.1 requires it, and declared every interest clause 5.2 covers; (d) you examined the full manuscript and the materials your report relies on, as attested in the report itself; and (e) your report does not infringe any third party's rights and does not contain defamatory or unlawful content.

9.2 · Integrity matters are referred, not decided. Where you encounter plagiarism, undisclosed duplicate publication, or signs that images or data have been manipulated, you must refer the matter to us and stop, and you must not write the allegation into a report that may be published. A report is not the place to make an allegation, and adjudicating one is not part of your engagement.

10. Term and termination

10.1 · Duration. This Agreement remains in effect for as long as you participate in the XSci reviewer program.

10.2 · Ending the relationship. Either party may end the relationship at any time. In-progress assignments should be completed or formally handed back before you withdraw. Handing back an assignment because clause 5.1 or clause 5.3 requires it is never a breach of this clause.

10.3 · Survival. Clauses that by their nature should continue — including Confidentiality (4), Review reports licence (6), Improvement authorization (7), and Representations (9) — survive termination.

10.4 · A report already delivered stands after termination, as does the assessment record naming it, and neither is withdrawn on account of your leaving the program.

11. Governing law

11.1 · Law and jurisdiction. This Agreement is governed by the laws of England and Wales, and the parties submit to the exclusive jurisdiction of its courts.

12. Contact

12.1 · Questions about this Agreement or your role as a reviewer: support@xsci.com

13. Updates

13.1 · Changes. We may update this Agreement to reflect changes in law or best practice. Reviewers will be notified of substantive changes and, where required, asked to re-accept before taking further assignments. An assignment you have already accepted is governed by the version in force when you accepted it, and the assessment record names that version.

13.2 · Acceptance. By applying to the reviewer program and accepting review assignments, you confirm that you have read, understood, and agree to this Reviewer Agreement.