可接受使用政策
本政策适用于本站账户、API 请求、输入和生成输出，是服务条款的一部分。使用时必须遵守适用法律、实际上游服务商的政策及第三方权利。AI 生成内容可能不准确；你须在使用和传播前审查，不能将其冒充为真实事实或未经 AI 处理的原始证据。
1. 禁止生成和传播的内容
- 色情及 NSFW：色情、露骨性行为、性剥削、非自愿私密影像及具有性暗示的违规内容。
- 暴力与血腥：鼓吹暴力、虐待、恐怖主义，或为了伤害、恐吓、剥削而生成的血腥内容。
- 仇恨：针对受保护群体的仇恨、歧视、去人格化、暴力威胁或煽动。
- 未成年人安全：任何儿童性虐待或性剥削材料，包括 AI 合成内容、性化未成年人、诱骗或伤害儿童的内容。
- 深度伪造与冒充：换脸、脱衣工具、伪造真人亲密场景、伪造真人从未说过或做过的言行，以及用于欺诈、误导或侵犯人格权的内容。
- 知识产权：未经授权复制、分发或使用受保护作品、商标及其他知识产权；不得伪称模型原厂身份或获得其背书。
2. 其他禁止行为
不得实施欺诈、钓鱼、垃圾信息、账户盗用、恶意软件、未授权入侵或有害攻击；不得规避安全过滤、权限、速率及额度限制，不得滥用试用、盗刷或操纵计费。未经本人同意或其他合法依据，不得上传或使用他人的照片、声音、身份资料或其他个人信息；即使取得同意，也不得使用本站制作上述禁止的深度伪造内容。
3. 风险处理与执行
本站可根据实际启用的安全配置、服务商的反馈、异常记录和用户举报调查风险，拒绝请求、限制调用、撤销密钥或暂停账户；法律要求时可保留相关证据或依法报告。模型厂商可能有其自身的安全控制，但不能据此推断本站已对每次提示词和每个生成输出进行完整审核。本政策不把尚未部署的审核能力描述为已上线功能。
生成能力和限制可能随安全措施、上游要求及法律要求调整。用户不得以切换模型、协议或账号等方式规避限制；不能因为请求被某个模型接受就认为内容符合本政策。
4. 举报与申诉
请将违规线索或申诉发送至 xinghuo@poviai.com，提供请求时间、订单或请求标识、问题说明及你有权提供的必要证据。不要发送完整 API 密钥或传播违法影像。我们会核实举报、采取适当措施，并在必要时请求补充信息；被误限制的用户可以提出复核。数据处理见隐私政策。
Acceptable Use Policy
This policy forms part of our Terms and applies to accounts, requests, inputs and generated outputs. Follow applicable law, actual upstream providers' policies and third-party rights. Review AI output before use and do not misrepresent it as authentic fact or unaltered evidence.
Prohibited content
- Pornography / NSFW: sexually explicit, exploitative, non-consensual intimate or prohibited sexually suggestive content.
- Violence / gore: content promoting violence, abuse or terrorism, or graphic material intended to harm, intimidate or exploit.
- Hate speech: protected-group hatred, discrimination, dehumanization, threats or incitement.
- Child safety / CSAM: any child sexual abuse or exploitation material, including synthetic content, sexualization, grooming or harm to minors.
- Deepfakes / impersonation: face swaps, nudification, fabricated intimate scenes or statements/actions of real people, and deceptive, fraudulent or rights-infringing impersonation.
- Copyright / trademark infringement: unauthorized copying, distribution or use of protected works, marks or other intellectual property, or false provider affiliation/endorsement.
Conduct and personal uploads
No fraud, phishing, spam, account theft, malware, unauthorized intrusion or harmful attacks. Do not bypass safety, access, rate or credit controls, abuse trials or manipulate payments. Do not upload another person's photos, voice, identity or personal information without consent or another lawful basis. Consent does not make prohibited deepfakes permissible.
Enforcement
We may investigate risks using actually enabled safety configurations, provider feedback, incident records and user reports, and reject requests, limit usage, revoke keys or suspend accounts. Evidence may be retained or reported where legally required. Provider controls do not establish that every prompt or output has undergone complete platform-side moderation; this policy does not claim undeployed controls are active. Capabilities may be restricted to meet safety, provider or legal requirements. Switching models or protocols must not be used to evade restrictions.
Reports and appeals
Email xinghuo@poviai.com with the time, relevant request/order reference, explanation and evidence you may lawfully share. Do not send complete keys or illegal imagery. We review reports, take appropriate action and may request further information. Users may appeal mistaken restrictions. See our Privacy Policy.