利用ポリシー（AUP）
Last updated: September 18, 2026
SegmentLens ("Platform", "we", "us") is committed to providing safe, lawful, and responsible AI content generation and image processing services. This Acceptable Use Policy ("Policy") applies to all content generated, uploaded, processed, published, transmitted, or stored through the Platform, and sets out our content standards, moderation practices, reporting procedures, and enforcement actions. All users, developers, and merchants who use the Platform must comply with this Policy.
This Policy complements our 利用規約. If there is a conflict regarding content standards, this Policy controls for content-related matters.
1. Scope
This Policy applies to all of the following content types on SegmentLens:
- Images: AI-generated or edited images, illustrations, segmentation masks, stickers, and related visual outputs
- Text: AI-generated text, chat replies, captions, and similar outputs
- User inputs (prompts): prompts, instructions, and uploaded source images submitted to the Platform
- Other forms supported by the Platform: including video segmentation, 3D-related assets, and API-submitted content where available
2. Content Standards and Prohibited Categories
The Platform expressly prohibits using our services to generate, upload, process, or distribute any of the following six categories of content:
- Sexual / NSFW content
- Violence / gore
- Hate speech
- Child sexual abuse material (CSAM) and other child-unsafe content
- Deepfakes / impersonation of real people
- Copyright / trademark infringement
The sections below further define these categories and related prohibited conduct.
2.1 Absolute Prohibitions — Zero Tolerance
The following content is subject to immediate enforcement. We reserve the right to report it to competent authorities:
- Child sexual exploitation material (CSAM) or any sexual content involving minors (anyone under 18), including fictional depictions
- Content that promotes or assists terrorism, extremism, or mass violence
- Guidance for producing weapons of mass destruction
- Content that incites genocide or ethnic hatred
- Non-consensual deepfake sexual content targeting real individuals
- Any other content that is illegal under applicable law
2.2 Strictly Restricted — Subject to Review
The following may be restricted or removed depending on context and applicable law:
- Graphic violence outside a clear educational, journalistic, or fictional creative context that does not glorify real-world harm
- Politically sensitive or misleading content that could deceive the public
- Medical, legal, or financial advice that does not meet applicable regulatory requirements
Note: Adult / NSFW content is not permitted on SegmentLens under any circumstances, including with age verification.
2.3 Generally Prohibited Conduct
- Generating or spreading false information or rumors intended to deceive
- Impersonating real individuals, brands, or institutions
- Infringing intellectual property, publicity, or privacy rights
- Presenting AI-generated content as human-created in order to deceive others
- Bulk generation of spam or fraudulent content
- Jailbreak prompts or other attempts to bypass Platform safety controls
3. Content Moderation
SegmentLens uses a multi-layer moderation system: automated screening + human review + periodic sampling.
3.1 Automated Moderation (Real Time)
Requests and generated content are screened by safety filters where technically applicable:
- Input filtering: keyword, semantic, and intent checks on prompts and related inputs
- Output filtering: safety assessment of generated or processed visual content
- Usage-pattern analysis: monitoring for abnormal API or product usage
- Safety model in use: OpenAI Moderation API (
omni-moderation-latest), applied to image uploads and related moderation checkpoints in our product flows
3.2 Human Review
- Content flagged by automation enters a human review queue
- User reports are reviewed by our trust and safety staff
- Review coverage: business-hours trust and safety review, with priority escalation for severe (L1) reports
- Reviewers complete content-safety training and are bound by confidentiality obligations
3.3 Periodic Sampling and Audit
- We perform random sampling of content and reports at least monthly
- We regularly evaluate and improve automated moderation accuracy
- We may engage third-party review where appropriate
3.4 Severity Levels
| Level | Description | Review | Enforcement |
|---|---|---|---|
| L1 Critical | Child safety, terrorism, WMD guidance, non-consensual sexual deepfakes | Auto-block + human confirmation | Immediate ban + report to authorities where required |
| L2 High risk | Clear severe policy violations | Auto-flag + priority review | Remove content + suspend account |
| L3 Medium risk | General policy violations | Human review | Warning + remove content |
| L4 Low risk | Minor violations | Report-triggered | Warning + require correction |
4. Reporting
4.1 When to Report
Please report:
- Content or behavior that violates this Policy
- Suspected child-safety, terrorism, or other severe violations
- AI impersonation or content that infringes your rights
- Attempts to bypass safety systems
4.2 Reporting Principles
- Confidentiality: we protect reporter identity to the extent permitted by law
- Neutrality: reports are reviewed independently and objectively
- Anti-abuse: malicious or knowingly false reports may be recorded and acted upon
5. How to Report
All reports, safety concerns, and appeals are handled through the same monitored inbox:
- Email: support@sam3.ai
- Contact form: お問い合わせフォーム
- Subject-line prefixes (please use so we can triage quickly):
[REPORT]— policy violations and abusive content[URGENT]— L1 / child safety / terrorism / other severe violations[APPEAL]— appeals of enforcement decisions[SAFETY]— other content-safety questions
Please include, where available: the content URL or history ID, approximate time, a short description of the issue, and relevant screenshots. Do not send illegal CSAM files; describe the location of the content so we can retrieve and preserve it securely.
6. Response Timelines
| Violation type | Initial response | Target resolution |
|---|---|---|
| L1 Critical | Within 2 hours | Within 24 hours |
| L2 High risk | Within 24 hours | Within 3 business days |
| L3 Medium risk | Within 3 business days | Within 7 business days |
| L4 Low risk | Within 5 business days | Within 15 business days |
If we need more time, we will notify the reporter when contact details are available and explain the reason.
Process: report received (acknowledgment) → severity grading (L1–L4) → content review (automated + human) → enforcement decision → action taken → outcome notice (where appropriate).
7. Enforcement Actions
| Action | Measure | Typical use |
|---|---|---|
| 1. Content warning | Written warning; require deletion or correction | First minor violation |
| 2. Content removal | Remove violating content from history/storage | Confirmed L3–L4 |
| 3. Feature restriction | Reduce quotas or model access | Repeated minor or single L3 |
| 4. Account suspension | Temporary suspension pending review | L2 or cumulative violations |
| 5. Permanent ban | Permanent termination of access | L1 or repeated L2 |
| 6. Legal referral | Report to law enforcement and cooperate with investigations | Suspected illegal L1 |
8. Appeals
If you disagree with an enforcement decision, you may appeal within 14 days of notice:
- Appeal channel: email support@sam3.ai with subject line starting with
[APPEAL] - Target review time: within 10 business days
- Review method: an independent reviewer who did not make the original decision
9. Transparency and Policy Changes
- Material changes to this Policy will be reflected by updating the "Last updated" date on this page
- Where practicable, we aim to provide advance notice of material changes
- Continued use of the Platform after an update constitutes acceptance of the revised Policy
10. Contact Us
- Reports, content safety, appeals, and general support: support@sam3.ai
- Contact form: お問い合わせフォーム
- Product: SegmentLens (https://sam3.ai)
This Policy is provided for operational transparency and does not constitute legal advice.