StartStep AI Complaints, Reports and Appeals Rules

Version: V1.11 Effective: This version takes effect on the date it is published in the Rules Centre.

Effective Date: This version takes effect on the date it is published in the Rules Centre.
Operator: Noviq Pte. Ltd.
UEN: 202607454K
Registered Office: 68 Circular Road, #02-01, Singapore 049422
Official Domain: startstep.ai
Customer Support: support@startstep.ai

Introduction

These StartStep AI Complaints, Reports and Appeals Rules (the “Process Rules”) explain how users and other persons may raise questions, complaints, content and safety reports, rights notices, privacy requests, security reports and appeals concerning StartStep AI, and how Noviq Pte. Ltd. (“Noviq,” “StartStep AI,” “we,” “us” or “our”) may receive, classify, verify, investigate, escalate and respond to them.

To reduce duplicate systems and operational burden, StartStep AI may use one intake system with internal routing to specialist processes. An in-product route should be used where available. If it is unavailable or the person has no account, the published email or web form may be used. We may use AI customer support for immediate guidance, collection of information, classification and summarisation. AI support cannot finally deny a valid statutory right or make a binding exception unless an authorised system flow or human representative confirms the result.

These Process Rules do not promise that every matter will receive human review, a different reviewer, a particular remedy or a fixed final resolution time, except where applicable law or a controlling Specific Service Rule requires it. We may prioritise matters by urgency, harm, legal duty, evidence, scope and operational capacity.

1. Scope, Terms and Related Rules

1.1 Scope

These Process Rules cover, as applicable:

  1. account registration, access, security, impersonation, restriction and termination;
  2. memberships, recurring payments, Compute Credits, official human services, orders, refunds and chargebacks;
  3. AI Output, file processing, workflows, Agents, model use, metering and technical faults;
  4. Community Content, reviews, harassment, illegal content, Synthetic Content and moderation;
  5. copyright, trade marks, privacy, publicity, reputation, confidentiality and other rights;
  6. Project contribution, Project acceptance, Project access and Project records;
  7. Project Host eligibility, Task Cards, Campaign Pages, participation, acceptance, submissions, metrics, attribution, rankings, review, commercial use, fixed fees, Campaign Rewards, Payables, tax, payout and suspension;
  8. personal-data rights, data-protection complaints and security incidents; and
  9. other matters for which the StartStep AI Rules identify this process.

Any person may report suspected illegal content, child-safety risk, security risk or an infringement of that person’s rights without registering for a StartStep AI account. Accepting such a report does not make an under-18 person eligible for an account or Service.

1.2 Definitions

In these Process Rules:

  1. “Enquiry” means a request for information, instructions or status that does not necessarily allege wrongdoing.
  2. “Complaint” means a statement that StartStep AI, a person acting through the Service or a relevant service outcome has adversely affected the complainant, together with a request for review, explanation, correction or remedy.
  3. “Report” means information alleging illegal, unsafe or Rules-violating content or conduct, whether or not the reporter is personally affected.
  4. “Rights Notice” means a notice alleging infringement of intellectual property, privacy, publicity, confidentiality, reputation or another enforceable right.
  5. “Privacy Request” means an exercise of a data-protection right, such as access, correction, deletion, objection, restriction, portability, withdrawal of consent or an applicable opt-out.
  6. “Privacy Complaint” means a complaint about how personal data has been processed; it is distinct from a Privacy Request even when submitted at the same time.
  7. “Appeal” means a request by a person materially affected by an eligible StartStep AI decision to reconsider that decision.
  8. “Security Report” means a report of compromised credentials, unauthorised access, malicious code, a vulnerability, data leakage, payment compromise or a comparable technical risk.
  9. “Request” means any of the above where the context permits.

We classify a Request according to its substance, not its title or the category selected by the requester. Marking a routine billing question “urgent,” describing a refund request as an illegal-content report or submitting the same issue in multiple categories does not change the applicable rules, evidence or priority.

1.3 Specialist processes prevail

The following control within their scope:

  1. refunds, cancellation and recurring payments: the Paid Services and Automatic Renewal Terms, Refund and After-Sales Policy, checkout disclosures and original payment-channel rules;
  2. Community Content and moderation: the Content and Community Rules;
  3. United States copyright notices and counter-notices: the DMCA Copyright Notice and Counter-Notice Policy;
  4. Privacy Requests and Privacy Complaints: the Global Privacy Notice, applicable regional privacy notice and mandatory data-protection law;
  5. Project contribution and Project Asset rights: the Project Contribution and AI Processing Rules;
  6. Project Host tasks, Campaigns and payments: the Independent Project Host Agreement, One-Off Host Task Service Fee and Settlement Rules, Limited-Time Host Operations Campaign Reward Rules, the accepted Task Card or joined Campaign Page and applicable regional schedule; and
  7. security vulnerability disclosure: any published security or coordinated-disclosure process.

Specific valid transaction, Project or task terms prevail over these general Process Rules for that specific matter, subject to the order in the International Rules Framework and mandatory law.

2. Intake Channels and Formation of a Case

2.1 In-product routes

Depending on the production feature, StartStep AI may provide one or more of:

  1. an order, subscription or refund control;
  2. a report button attached to content or an account;
  3. an appeal control in an enforcement notice or Project, Task or Campaign page;
  4. a privacy-rights request form or account-deletion control;
  5. an AI support, help-centre or customer-service interface; and
  6. a security, rights or legal-notice form.

We are not required to build a separate case-management system for every category. Different interfaces may create records in the same underlying system.

2.2 Published remote channels

Where an in-product route cannot be used, the following channels apply:

  1. general support: support@startstep.ai;
  2. content and safety: support@startstep.ai;
  3. privacy: privacy@startstep.ai;
  4. security: security@startstep.ai;
  5. legal notices other than DMCA: legal@startstep.ai; and
  6. DMCA: the agent and channel specified in the DMCA Policy.

We may later route several published addresses to a unified case system. An address may accept messages at any time, but this does not promise 24-hour human staffing or immediate human response.

2.3 Formal and informal contact

A Request is formally submitted when it reaches a designated product route or published address in a reasonably readable form. A public post, app-store review, social-media comment, message to an unofficial account, communication with a Project Host or private contact with an assigned Done-for-You worker is not a formal refund request, Rights Notice, Privacy Request, Appeal or legal notice unless the applicable Rules expressly say otherwise.

We may nevertheless act on an apparent serious risk that we independently discover through an informal channel.

2.4 Case reference and consolidation

We may issue a case reference, associate related messages with an existing case, split unrelated issues or consolidate duplicate or coordinated submissions. Failure to receive a case number does not by itself invalidate a Request that applicable law treats as received.

2.5 Legally effective requests

International Terms Section 15.2 applies to receipt, internal routing, verification and statutory clocks. A legally effective request is not invalid merely because it used another sufficient route. Mandatory destination requirements, including those applicable to formal DMCA notices and legal process, remain.

3. AI Support and Human Escalation

3.1 Permitted AI support functions

AI support may:

  1. answer routine questions using published information;
  2. help select a category and identify likely supporting materials;
  3. summarise messages and create or update a case;
  4. retrieve an available order, account or case status after authentication; and
  5. suggest self-service steps that the user remains free to accept or reject.

3.2 Limits of AI support

Unless confirmed through an authorised transaction flow or by an authorised human, AI support cannot:

  1. waive or amend the StartStep AI Rules;
  2. promise or finally reject a refund, compensation or payment;
  3. restore or terminate an account or Project Host relationship;
  4. decide a complex Rights Notice, Privacy Complaint or statutory entitlement;
  5. accept service of legal process; or
  6. create a binding admission of liability.

3.3 Human escalation

We may route a matter to a human process when it involves material billing, identity, account ownership, privacy, child safety, physical danger, intellectual property, account termination, Project Host cash settlement, a legal right, a competent-authority request or an issue the AI cannot reliably handle.

Human escalation is subject to reasonable verification and available support operations. Unless law requires otherwise, it does not promise a live conversation, immediate response, a different reviewer or unlimited levels of review.

4. Information, Evidence and Verification

4.1 Information to provide

A requester should provide the applicable parts of the following:

  1. reliable contact information;
  2. account identifier and registered email or telephone number;
  3. order number, payment channel, date, currency and amount;
  4. content URL or identifier, account, Project, AI task, workflow, Agent, Host Task ID, Campaign ID, statement or payment event;
  5. relevant dates and a concise chronology;
  6. the specific issue, basis and requested outcome; and
  7. original files, screenshots, correspondence, authority documents or other reasonably relevant evidence.

Do not provide passwords, one-time codes, private keys, full card-security codes, unnecessary identity-document data or unrelated sensitive information. Noviq will not request those secrets to process an ordinary complaint.

4.2 Verification

To prevent account theft, fraudulent refunds, unauthorised deletion or disclosure and false rights claims, we may require login verification, one-time-code verification, historic account or transaction information, identity-verification results, a redacted identity document, authority evidence or another proportionate method.

An agent or representative must show authority. An organisation must show the requester’s authority to act for it. A parent, guardian, executor, attorney or other legal representative must provide the relationship and authority required by applicable law.

Requesters may redact information irrelevant to verification. If proportionate verification cannot be completed, we may pause or refuse an action that could affect another person, disclose protected information or transfer value, while continuing to address any independently apparent safety or legal risk.

4.3 Incomplete or unclear Requests

We may ask for clarification or additional material where a Request cannot be located, its requested outcome is unclear, information conflicts or verification is incomplete. A statutory period is calculated and may be paused, restarted or extended only as the applicable law permits.

Where no mandatory period applies, we may close a case if the requester does not provide reasonably requested information, indicates that no further action is wanted, cannot be contacted, or the matter cannot reasonably be verified. A later submission with new information may be opened as a new case but does not automatically restore an expired programme, internal appeal period or deleted data.

4.4 Truthfulness and preservation

The requester must act in good faith, preserve relevant evidence and provide information that is materially accurate and not misleading. A report need not prove the final legal case, but fabricated, altered or knowingly incomplete evidence may result in rejection, account action or legal responsibility.

5. Triage, Investigation and Response

5.1 General process

We may:

  1. receive and record the Request;
  2. classify it by type, urgency and controlling rule;
  3. authenticate the requester and locate relevant records;
  4. take an interim measure or notify an affected party or provider;
  5. investigate using reasonably available evidence;
  6. grant, partly grant, deny, redirect, close or continue the matter;
  7. communicate an outcome or appropriate progress information; and
  8. retain necessary records.

Steps may be combined, reordered or omitted based on the issue. We are not required to conduct litigation-style discovery, obtain evidence for a party, decide facts beyond our reasonable capability or provide legal advice.

5.2 Priority

We may prioritise according to imminent physical danger, child safety, active fraud, security compromise, legal deadline, scale, number of affected persons, continuing harm, strength of the available signal, ability to reduce harm and operational capacity.

A user’s requested priority or “urgent” label is relevant but not controlling. We do not guarantee immediate human action. For imminent danger, crime in progress, urgent medical need or immediate payment blocking, contact local emergency services, law enforcement, a bank, payment provider or other competent body as well as StartStep AI.

5.3 Interim measures

Pending review, we may restrict content, links, accounts, transactions, withdrawals, Project access, Project Host payout, automation, credentials or other functionality where reasonably necessary to prevent continuing harm, preserve evidence or comply with law. An interim measure is not a final finding and an Appeal does not automatically suspend it.

5.4 Timing

Unless a mandatory legal period, a controlling Specific Service Rule or a specific written case notice provides otherwise, we do not promise a fixed acknowledgement, progress-update or final-resolution period. We will process Requests within a period reasonable for their nature, completeness, risk, technical complexity, third-party dependency and operational capacity.

We may require more time for multiple parties or accounts, provider or app-store action, payment investigation, data restoration, security forensics, ownership disputes, expert review, legal process, tax or payout verification, fraud analysis or a material change in facts. We may provide progress information without promising a final date.

5.5 Communications and delivery

We may respond through the original route, account notification, order or case page, registered email, telephone, SMS or another reasonable channel. A response is effective when reasonably sent, subject to mandatory delivery rules. The requester must maintain accurate contact details and review spam folders and in-product notifications.

We may provide the response in English and, where available, another language. The controlling language rules in the International Terms continue to apply.

6. Content, Safety and Illegal-Content Reports

6.1 Who may report

Any person may report content or conduct believed to be illegal, unsafe or contrary to the Content and Community Rules. A reporter need not be the person directly affected, although a Rights Notice requires appropriate authority.

6.2 Report content

A useful report identifies the content or account, exact location, date, alleged category, reason, context, affected person and supporting information. A report of suspected illegal content should explain why the reporter believes it is illegal where reasonably possible.

Do not transmit child sexual abuse material or non-consensual intimate imagery by ordinary email. Provide a URL or content identifier and use any secure instructions. If we cannot locate the item, we may request more information or close the report while retaining authority to act on independently identified risk.

6.3 Assessment and measures

We may consider the report, content, context, reach, user history, automated signals, third-party information and applicable law. We may label, restrict, de-index, disable interaction, remove, preserve, limit an account or take another measure under the Content and Community Rules.

A report does not automatically require removal. Closing or declining a report does not certify that content is true, lawful, non-infringing or suitable. Moderation is a risk-management decision, not a court or regulatory ruling.

6.4 Feedback and confidentiality

We provide acknowledgement, outcome, reasons, review or other information to the extent required by applicable law and the Rules. A reporter who is not directly affected may receive only limited status information. We may withhold another person’s personal data, reporter identity, complete evidence, security methods, moderation thresholds, privileged information and details that would enable evasion or increase harm.

6.5 Non-consensual intimate imagery and digital forgeries

To request removal of an intimate image or video shared without consent, including an AI-generated or altered depiction, email support@startstep.ai or use the content report control. An account or purchase is not required. Identify the material and location sufficiently for us to find it, provide a contact method, a physical or electronic signature, and a brief good-faith statement that the depiction was published without the depicted person's consent and relevant information supporting that statement. An authorised representative may act for the person. Do not create or send additional copies of illegal imagery merely to make a report; use a location reference where possible.

Where the U.S. TAKE IT DOWN Act applies to the relevant platform and request, Noviq will remove the identified depiction as soon as possible and no later than 48 hours after receiving a valid request, and make reasonable efforts to identify and remove known identical copies within that period. The period runs continuously, including weekends. A DMCA counter-notice waiting period or ordinary appeal does not delay this process. Other applicable urgent removal, preservation and reporting duties remain effective. We limit disclosure of victim information to lawful, necessary purposes.

7. AI Feature Complaints

7.1 Covered matters

A person may report or complain about apparently illegal or seriously harmful Output, deceptive impersonation, missing synthetic-media labels, material metering errors, repeated charges for the same task, unauthorised Agent action, file-processing failure or another apparent breach of the AI Notice.

7.2 Supporting information

Provide the task identifier, date, feature, model label if displayed, relevant Input and Output, files, error state, Credit record, expected behaviour and actual impact. Do not disclose another person’s confidential or sensitive data unless necessary and lawful.

7.3 Review and available action

We may review relevant task logs, route records, Credit events, error codes, safety labels and versions. We do not have to disclose system prompts, chain-of-thought or internal reasoning, model weights, security controls, confidential provider information or details that would enable circumvention.

Depending on the facts and applicable Rules, we may explain a limitation, correct manifest metering, restore eligible Credits, regenerate, change routing, add a warning or label, restrict an Output, fix a function or take no further action. Dissatisfaction with style, usefulness or subjective quality does not automatically establish a refund or compensation right.

8. Accounts, Security, Billing and Paid Services

8.1 Account ownership and access

Account recovery, impersonation, suspected compromise and binding changes must be submitted from a trusted channel and pass proportionate account-control checks. While ownership or risk is unresolved, we may restrict login, payment, publication, Project submission, personal-data access, automation or Project Host payout.

The user should promptly secure credentials, connected accounts and devices. Noviq will not ask for a password or one-time code through an ordinary support exchange.

8.2 Unauthorised transactions

For a suspected unauthorised charge, promptly contact the bank, app store or responsible payment provider and provide StartStep AI with the order, time, amount, currency and channel. We may review available account, device, authorisation, benefit-use and risk records and may suspend the disputed benefit or related high-risk activity.

Noviq can act only within its control and does not replace a financial institution, app store, card network or court in determining payment liability.

8.3 Refunds, cancellations and recurring payments

Eligibility, deadlines, used-value adjustments, cancellation, refund route and timing are governed by the Refund and After-Sales Policy, Paid Services and Automatic Renewal Terms, checkout disclosure, applicable Regional Supplement, payment provider and mandatory law.

Apple, Google or another original payment channel may control subscription cancellation or refund review. A support case does not itself cancel a subscription unless the relevant flow confirms cancellation. We may provide available assistance but do not control a third party’s processing or bank-posting time.

8.4 Chargebacks

A lawful chargeback right is not restricted. A chargeback may nevertheless cause the disputed order, Credits, membership, Project Host attribution or payout to be suspended or reversed while the matter is reviewed. Fraudulent, duplicative or bad-faith chargebacks may lead to account restriction and recovery action to the extent permitted by law.

No person may recover twice for the same loss through refund, replacement, chargeback, insurance, settlement or damages.

8.5 Official human services

A complaint concerning an Official Done-for-You Service should identify the signed services agreement or quote, accepted scope, required materials, deliverable, delivery status, alleged material non-conformity and supporting information. Personal dissatisfaction, a later preference change or failure to achieve an unstated business result does not by itself establish non-performance. The signed services agreement and its quote, and the applicable acceptance, cancellation and refund rules, control. Mandatory consumer rights are not affected.

9. Privacy Requests, Privacy Complaints and Data Security

9.1 Separate classification

A Privacy Request and a Privacy Complaint are legally and operationally distinct even when contained in one message. We may split them into separate cases and apply different verification, timing, exemptions and appeal rules.

9.2 Privacy Requests

Eligible persons may exercise the rights described in the applicable Privacy Documents. We may verify identity and authority, clarify scope, apply legal exceptions, charge a fee only where law permits, and retain information required for transactions, tax, fraud, security, legal claims or other lawful purposes.

Account deletion or withdrawal of consent does not automatically reverse a completed Project Asset rights transfer, erase non-personal data, cancel a subscription, create a refund, remove another user’s records or delete information that law permits or requires us to retain.

9.3 Privacy Complaints

A person may complain that personal data was processed unlawfully, inaccurately, insecurely or contrary to a Privacy Document. We will provide a clear route, investigate appropriately, keep the person reasonably informed and communicate an outcome within the periods required by applicable law.

9.4 Security Reports

Report suspected data leakage, unauthorised access, exposed credentials, malicious code or a vulnerability through the security channel. Provide the affected feature, date, risk and safe reproduction information. Do not access additional data, disrupt systems, demand payment, publicly disclose an unremediated vulnerability or transmit active secrets or large sensitive datasets through ordinary email.

We may restrict access, invalidate credentials, isolate systems, preserve evidence, investigate, notify affected persons and report to authorities as required by law. A good-faith report does not create a right to a bounty, employment, confidentiality agreement or particular response unless a published programme says so.

10. Intellectual Property and Other Rights Notices

10.1 DMCA matters

United States copyright notices and counter-notices within the DMCA Policy must be submitted to the designated DMCA agent using the required content. A general support message does not substitute for a legally compliant DMCA notice, although we may redirect it.

10.2 Other Rights Notices

A notice concerning trade marks, copyright outside the DMCA process, privacy, publicity, likeness, voice, reputation, confidentiality or another right should include:

  1. the claimant’s identity and reliable contact information;
  2. authority if submitted by a representative;
  3. the right and reasonable supporting evidence;
  4. the exact location of the challenged material;
  5. the relevant facts and requested measure; and
  6. a good-faith statement that the information is accurate and the claimant is authorised.

10.3 Forwarding and interim action

We may forward relevant portions to the affected user, request more material, restrict content temporarily or decline an unsupported notice. Taking a measure does not admit liability or finally determine infringement. Complex disputes concerning ownership, licence, fair use or damages may require court, arbitration or regulator action.

10.4 False or abusive notices

A person submitting a materially false or bad-faith notice may be responsible under applicable law. We may limit frequency, require stronger verification, restore affected content, take account action or seek recovery where permitted, without obstructing legitimate rights enforcement.

11. Appeals from StartStep AI Decisions

11.1 Eligible decisions

Subject to the controlling Rules, an Appeal may be available for:

  1. rejection, labelling, visibility restriction or removal of content;
  2. restriction of reviews, recommendations, automation or community features;
  3. account suspension, feature restriction or termination;
  4. a manifest Compute Credit record or metering decision;
  5. Project submission status, contribution record or eligible Credit reward;
  6. Project Host eligibility, Task offer, Campaign Page, joining, acceptance, submission, metric, attribution, observation adjustment, rank, human confirmation, commercial-use, fixed-fee, Campaign Reward, hold, Payable or settlement record; and
  7. another decision whose notice expressly offers an Appeal.

An Appeal is not the correct route for a new refund request, DMCA counter-notice, statutory Privacy Request, emergency report or disagreement with an independent third-party decision.

11.2 Appeal period

The period shown in the decision notice or Specific Service Rule applies. If none is shown, an internal Appeal should normally be submitted within 14 calendar days after the affected person receives or reasonably becomes aware of the decision.

This internal period does not shorten a limitation period or prevent a non-waivable consumer, privacy, payment, safety, intellectual-property or external legal remedy. We may consider a late Appeal involving a manifest system error, serious safety issue, new material evidence or another matter we are legally required to address.

11.3 Appeal material

Identify the decision or case and explain the alleged error, relevant context, source, rights, evidence, corrective action and requested outcome. An Appeal should raise all reasonably available material points; merely repeating disagreement without supporting information may be insufficient.

11.4 Review and outcome

We may use automated, human or combined review as permitted by law. Human review or explanation will be provided where mandatory law requires it. Unless law requires otherwise, review need not be performed by a different individual, and we do not promise more than one internal Appeal on materially unchanged facts.

We may uphold, vary, reverse or replace a decision, correct a record, restore an eligible benefit, permit resubmission or take another appropriate action. Restoration does not guarantee the original publication time, ranking, engagement, programme opportunity, Project traffic or another state that cannot reasonably be recreated.

11.5 No automatic stay

An Appeal does not automatically stay a content, account, payment, security or safety measure. We may narrow or suspend the measure while reviewing if we consider it appropriate.

12. Project and Project Host Disputes

12.1 Project contribution records

A Project Contributor may request correction of a Project snapshot, agreement version, confirmation event, source declaration, acceptance status, contributor display or eligible Compute Credit reward. A completed transfer or licence of Project Asset rights is not automatically rescinded by a complaint, account deletion or privacy request; applicable law and the Project rules govern any challenge to validity.

12.2 Project Host task records

A Project Host disputing a Task offer, locked Task Card, Campaign Page, participation record, submitted version, metric, attribution, observation adjustment, provisional or final rank, review outcome, human confirmation, deemed acceptance, commercial-use event, fixed fee, Campaign Reward, Payable, tax item, payout hold or payment status should identify the Host, Project, Task or Campaign ID, statement or payment reference, applicable version and supporting evidence.

An unaccepted Task offer, unjoined Campaign, provisional metric, progress bar, temporary rank or estimated opportunity is not an amount due. A fixed Gross Service Fee becomes payable only through the applicable Task Payable Event. A fixed Campaign Reward becomes payable only after Campaign close, the disclosed observation and integrity checks and authorised human confirmation. We may correct duplicate payments, obvious errors, fraud, collusion, related-account activity, false materials, infringement, unlawful Deliverables or other legally recoverable amounts, but an ordinary User refund or internal business change does not by itself reverse an accepted fixed fee or a Campaign Reward already finally confirmed after the applicable observation period.

12.3 Payout and tax dependencies

We may hold a payout while required identity, authority, bank, sanctions, tax or provider review is incomplete. Noviq can verify its task and payment records but does not replace a bank, tax authority or independent payout or tax provider in deciding matters outside Noviq’s control.

12.4 No user-Host transaction dispute

A Project Host is not the seller of a User’s StartStep AI purchase and is not responsible for the User’s refund merely because the Host supported or promoted a Project. User purchase complaints are directed to Noviq or the merchant identified at checkout.

13. Records, Personal Data and Confidentiality

13.1 Records

We may retain cases, messages, submitted documents, verification results, orders, Project snapshots, task and model logs, content versions, moderation actions, delivery records and outcomes as reasonably necessary for the Request, legal compliance, security, audit, fraud prevention and disputes, subject to the Privacy Documents.

Ordinary business records may be used as initial evidence in our internal review. A requester may submit contrary evidence. Our records do not displace legally admissible evidence or bind a court or regulator.

13.2 Sharing for case handling

We may share necessary information with an affected user, rights holder, app store, payment or identity provider, AI or cloud provider, Project Host service provider, professional adviser, insurer, safety organisation or authority where authorised or required by law and reasonably needed to process the matter.

Requesters should redact irrelevant personal data, confidential information and secrets. We may redact, limit access to or refuse to forward irrelevant or disproportionately sensitive material.

13.3 Information we may withhold

Subject to non-waivable rights, we may withhold:

  1. a reporter’s non-public identity and contact details;
  2. another person’s account, order, payment, Input, Output, tax or personal data;
  3. internal moderation guidance, model information, fraud thresholds and security controls;
  4. provider contracts, pricing, architecture, vulnerabilities and trade secrets;
  5. privileged material and confidential investigation or authority information; and
  6. information whose disclosure would facilitate evasion or materially increase harm.

14. Duplicate, False and Abusive Requests

14.1 Restrictions

To the extent permitted by law, we may consolidate, limit, close or require additional verification for Requests that are:

  1. materially duplicative and unsupported by new information;
  2. impossible to locate, irrelevant, automated spam or advertising;
  3. submitted using impersonation, forged orders, authority or evidence;
  4. used for extortion, harassment, retaliation, censorship of honest criticism or disruption;
  5. accompanied by threats or abuse of support staff, reporters or affected persons;
  6. generated at scale to impose disproportionate operational burden; or
  7. manifestly unfounded or excessive under an applicable law.

14.2 Protected good-faith activity

We will not apply this section merely because a person makes a good-faith complaint, gives a negative review, exercises a consumer or privacy right, reports suspected illegality or security risk, contacts a regulator or seeks a lawful remedy. A person may be wrong without acting abusively.

15. Outcomes, Remedies and External Routes

15.1 Possible outcomes

Within our authority and subject to the controlling Rules, we may:

  1. provide information, instructions or a risk warning;
  2. request verification, evidence, correction or mitigation;
  3. correct a record, reperform a controlled step or restore an eligible benefit;
  4. grant, partly grant or deny a refund or route it to the original channel;
  5. label, restrict, disable, preserve or remove content;
  6. restrict, restore, suspend or terminate an account, Project, Project Host status or feature;
  7. correct a Project, task, attribution, score, tier or settlement record;
  8. forward the matter to another person or provider;
  9. continue investigation, preserve evidence or report to an authority; or
  10. take another measure permitted by the Rules and law.

A goodwill Credit, refund, temporary benefit or assistance is not an admission of liability or a promise of identical treatment in another matter.

15.2 Third-party limits

We may assist with an app store, bank, payment provider, identity provider, model provider, cloud provider or tax and payout provider, but do not control its independent decision or processing time. The requester may need to use that party’s process.

15.3 External remedies

These internal processes do not prevent any person from contacting a competent regulator, consumer body, privacy authority, law-enforcement agency, court, arbitration body or other lawful forum. Internal discussion is not a mandatory precondition to emergency, injunctive or other action where applicable law provides otherwise.

Noviq does not promise that an external body has jurisdiction, will accept a Request or will award a particular remedy. The requester remains responsible for external filing requirements and limitation periods.

15.4 Liability and dispute terms

The International Terms and applicable Regional Supplement govern disclaimers, liability, governing law and dispute resolution. These Process Rules do not create a new damages remedy, admission standard or separate liability cap.

16. Territory-Specific Requirements

16.1 General

Mandatory local periods, notices, appeal rights and regulator information apply within their scope and prevail over a general provision here. A statutory process is not replaced by re-labelling a matter as a general customer-support case.

16.2 Singapore

For an eligible Singapore personal-data access or correction request, Noviq will respond as soon as reasonably possible and, if unable to provide access or make the correction within 30 calendar days, will provide the written timing information required by the Personal Data Protection Act 2012 and applicable regulations. Other Singapore Privacy Requests, complaints and review rights are governed by the Privacy Documents and applicable law.

Where a valid Singapore authority direction or mandatory online-safety process applies, we may prioritise and handle the matter under that legal process rather than the ordinary sequence here.

16.3 United Kingdom

For a United Kingdom data-protection complaint, Noviq will provide a clear channel, acknowledge receipt within 30 calendar days or any different period then required by law, take appropriate steps to investigate, keep the complainant appropriately informed and communicate an outcome without unjustifiable or excessive delay. A separate data-subject rights request follows its own statutory period.

Where a StartStep AI feature is within the United Kingdom Online Safety Act 2023, eligible users and affected persons may use the reporting and complaints routes required for illegal content, content-reporting decisions, terms enforcement and other matters within scope. We will provide the information and review required by that law; the procedure does not create broader duties for features outside its scope.

16.4 United States

United States state Privacy Requests and statutory appeals are governed by the United States State Privacy Notice, including the response period, authentication, authorised-agent and attorney-general information applicable to an eligible requester. No privacy appeal is promised where applicable law does not require one, although we may reconsider voluntarily.

DMCA notices and counter-notices follow the DMCA Policy, including its declarations, agent, forwarding and timing rules. A general Appeal under section 11 does not replace a DMCA counter-notice.

17. Changes and Contact

17.1 Changes

We may update these Process Rules for changes in law, channels, product features, risks and operations. Material changes will receive the notice and renewed acceptance required by law or the International Terms. Channel, form, category, security and workflow changes that do not reduce a non-waivable right may take effect when posted or implemented.

17.2 Contact directory

General support: support@startstep.ai
Request and report portal: [REQUEST AND REPORT PORTAL URL — TO BE COMPLETED BEFORE LAUNCH]
Content and safety: support@startstep.ai
Privacy: privacy@startstep.ai
Security: security@startstep.ai
Legal notices: legal@startstep.ai
DMCA: use the designated agent and channel in the DMCA Policy.

Postal address: Noviq Pte. Ltd., 68 Circular Road, #02-01, Singapore 049422.

Include only the information reasonably necessary for the Request. Do not send passwords, one-time codes, private keys, full payment credentials, CSAM or non-consensual intimate imagery.

Appendix A — Routing Summary

MatterPreferred routeControlling document
General product questionAI support/help centre/general supportInternational Terms and relevant feature rule
Order, refund or cancellationOrder/subscription page or general supportPaid Services Terms, Refund Policy, checkout and channel rules
Content or safety reportContent report button or content/safety channelContent and Community Rules
Content or account AppealLink in decision notice or appeal formDecision notice, these Process Rules and relevant Specific Rule
US copyrightDMCA agentDMCA Policy
Other Rights NoticeLegal-notice channelThese Process Rules and applicable law
Privacy RequestPrivacy form/emailGlobal and regional Privacy Documents
Privacy ComplaintPrivacy channel marked “Privacy Complaint”Applicable Privacy Documents and law
Security incident or vulnerabilitySecurity channelSecurity process and applicable law
Project contribution recordProject page or general supportProject Contribution and AI Processing Rules
Project Host Task, Campaign or payoutHost Task/Campaign/settlement page or Host supportProject Host agreement, Task and Campaign rules, applicable page and regional schedule
EmergencyLocal emergency service or authority, plus report if usefulApplicable law; StartStep AI is not an emergency service

Back to Rules