Community guidelines & acceptable use

Operated by Kinetic Helix LLC, a Florida limited liability company (doing business as “Sealed Rose,” “we,” “us,” “our,” or the “Company”). Contact: [email protected].

Sealed Rose is dedicated to providing high-integrity, autonomous artificial intelligence (AI) forensics and synthetic media verification. These Guidelines establish the standards of conduct and acceptable use expected of all individuals, researchers, and enterprises utilizing our public detection tools, developer APIs, and reverse face search utilities.

1. Who this applies to

These Guidelines apply to everyone who accesses Sealed Rose—including anonymous web visitors, registered account holders, and developer API consumers. By accessing our forensics engines, you agree to comply with these rules.

2. Core principles

• Probability, not proof: We give probability-based assessments, not proof. Our neural networks, spatial analyses, and spectral classifiers measure probabilistic indicators of synthetic manipulation to help individuals and organizations triage suspicious content.
• Likeness defense & consent: Every person has an inherent right to prevent the unauthorized, weaponized synthetic cloning of their face, voice, and personal identity.
• Ephemeral processing: We treat user media with utmost confidentiality. Media submitted for verification is analyzed ephemerally in volatile memory and deleted immediately after your scan. We never train public foundation models on your uploads.
• Uncompromising child protection: We enforce strict, zero-tolerance defenses against any form of child sexual abuse material or exploitation imagery.

3. Permitted & encouraged use

Sealed Rose is engineered for lawful, constructive media verification, including:

• Catfish & impersonation defense: Analyzing suspicious photos, videos, or voice notes received on dating apps, social networks, or messaging platforms to verify whether the sender is a real person or a synthetic deepfake persona.
• Likeness & face search: Searching indexed web media using facial vector embeddings to identify unauthorized deepfake depictions or stolen likenesses.
• Journalistic & OSINT research: Fact-checking viral video clips, generated imagery, or suspicious voice recordings to combat synthetic disinformation.
• Enterprise trust & safety: Integrating our developer APIs (/api-keys) into platform onboarding and content pipelines to screen for synthetic fraud and bot automation.

4. Prohibited conduct

The following activities are strictly prohibited across all Sealed Rose services:

(a) Minors and child sexual abuse material (CSAM) — zero tolerance. Submitting, scanning, or processing any visual depiction of child sexual abuse material or the sexualization of minors is strictly prohibited. Every image and video frame is hashed against Microsoft PhotoDNA. If suspected CSAM is detected, processing stops and we report it to the NCMEC CyberTipline as required by 18 U.S.C. § 2258A. The reported material is preserved in restricted, encrypted storage for the period the law requires (under 18 U.S.C. § 2258A(h)), then deleted.

(b) Non-consensual deepfakes & intimate imagery (NCII). Uploading, analyzing, or transmitting non-consensual sexually explicit deepfakes or unauthorized intimate imagery is strictly prohibited under the federal TAKE IT DOWN Act (enacted into law in May 2025, with platform notice-and-removal mandates taking effect in May 2026) and applicable state laws. If an unauthorized deepfake or intimate image of you has been submitted or indexed, you may submit an expedited removal request through our Non-Consensual Imagery Removal Form. In alignment with statutory notice-and-removal standards, valid requests are acted upon within 48 hours, and perceptual hashes are registered to prevent re-processing.

(c) Extortion, blackmail & harassment. Utilizing our forensic reports or reverse search tools to harass, dox, extort, or stalk any individual is strictly prohibited.

(d) Adversarial model evasion testing. Submitting iterative media variants solely to benchmark or train generative AI models to evade deepfake detection, or attempting to reverse engineer our proprietary neural scoring pipelines, is prohibited.

(e) Automated endpoint abuse & scraping. Bypassing rate limits through proxy cycling, botnets, automated scraping, or denial-of-service (DDoS) attacks will result in immediate IP blocking, credit forfeiture, and legal referral.

5. Reporting & takedown requests

We provide rapid response mechanisms for safety and rights protection:

• Non-consensual deepfakes / NCII: Submit via our no-login Removal Form under our expedited 48-hour SLA in accordance with the TAKE IT DOWN Act (enacted May 2025, platform notice-and-removal effective May 2026).
• Copyright & DMCA: Review our DMCA Policy to submit a formal notice to our Designated Agent.
• Abuse & security reports: Email our security team directly at [email protected].

6. Enforcement & law enforcement cooperation

Violations of these Guidelines result in swift enforcement, including warnings, immediate content quarantine, temporary suspension, or permanent account and IP bans.

Sealed Rose complies with valid legal process, including subpoenas, warrants, and court orders issued by competent judicial authorities. In instances of CSAM or imminent threats to human safety, we proactively notify law enforcement agencies as required or permitted by federal law.

7. Updates to these guidelines

We may update these Guidelines periodically as generative AI technologies and regulatory frameworks evolve. The effective date at the top of this document indicates the most recent revision.