Acceptable Use Policy

This Acceptable Use Policy applies to Coul accounts, uploads, prompts, references, linked sources, generated and edited outputs, profile connections, analytics, nodes and workflows, private, unlisted and public templates, APIs, exports, schedules, agents, automations, and publishing attempts.

You must not request, attempt, facilitate, conceal, coordinate, or enable prohibited conduct, even if no output is generated or no post is published. You are responsible for people and systems using your account or workspace within your control and for reviewing every output before use.

Destination-platform rules, laws, professional duties, ad rules, and source licenses also apply and may be stricter. Coul's ability to generate, analyze, schedule, or technically transmit something does not mean the use is permitted, accurate, safe, or platform-compliant.

Use Coul only for a lawful purpose and with every authorization, consent, license, account permission, and professional qualification the activity requires. Do not misstate identity, authority, affiliation, sponsorship, evidence, source, approval, or the origin of content.

Do not use Coul to violate sanctions, export controls, court orders, platform access rules, advertising restrictions, privacy or publicity rights, confidentiality, consumer-protection law, or another person's legal rights. Do not help someone evade a restriction that would apply if they acted directly.

Do not sell, sublicense, rent, resell, or provide unauthorized access to Coul, a connected social account, provider credential, dataset, model endpoint, or product output in a way that circumvents plan, platform, security, or usage controls.

Do not upload, reproduce, re-post, distribute, closely imitate protected expression, remove watermarks or rights-management information, counterfeit branding, create trademark confusion, or use another person's music, footage, image, voice, likeness, private material, or confidential information without the rights required for the intended use.

Coul's competitor feed may be used for public facts and metrics, themes, topics, general techniques, unprotected formats, and original ideation. Do not submit a competitor video to the virality analyzer or editor, reproduce its script or caption, trace its distinctive expression, reupload its media, impersonate its creator, or imply endorsement or affiliation.

Fair use and fair dealing depend on jurisdiction and facts. Attribution, a link, public access, an internal reference-only label, or an AI transformation does not by itself grant permission or cure infringement. Use the Copyright/DMCA Policy for takedowns and counter-notices.

Do not create or operate deceptive impersonation accounts; forge endorsements, credentials, evidence, or communications; clone or simulate a person's voice or likeness without appropriate authorization; or materially manipulate media to falsely portray a real person's words, actions, beliefs, condition, or involvement.

Do not remove a required provenance marker, watermark, metadata record, platform label, or visible synthetic-media disclosure. Apply clear context when a reasonable viewer could mistake materially generated or altered media for authentic, and comply with the AI Transparency notice and destination rules.

Obvious parody, satire, fiction, documentary, reenactment, and artistic uses may be allowed when lawful, clearly contextualized, and not used for fraud, harassment, sexual exploitation, voter deception, or misleading affiliation. Do not use faces or voices to infer identity, emotion, criminality, health, sexuality, politics, or another sensitive trait.

Do not dox, stalk, threaten, steal an identity, expose credentials, harvest personal data unlawfully, record someone unlawfully, or publish precise location, contact, financial, health, intimate, or other information in a way likely to create serious risk. Do not use Coul for unlawful surveillance, facial or voice identification, or sensitive-trait profiling.

Never create, request, upload, store, share, threaten to share, or facilitate child sexual abuse or exploitation material, sexualized depictions of minors including synthetic depictions, grooming, sextortion, trafficking, solicitation, or instructions that facilitate abuse. Never create or distribute real or synthetic non-consensual intimate imagery or sexual deepfakes.

Lawful family, educational, news, documentary, or awareness content involving minors is not prohibited merely because a child appears, but users must have appropriate rights, protect privacy and safety, avoid sexualization or exploitation, and follow age, platform, and regional rules.

Do not make credible threats; incite abuse; coordinate targeted or repeated harassment; engage in sexual harassment, blackmail, coercion, humiliation campaigns, doxxing, or dogpiling; dehumanize people; or promote exclusion, violence, or severe harm based on actual or perceived race, ethnicity, nationality, caste, religion, sex, gender, gender identity, sexual orientation, disability, serious disease, age, or another comparable legally protected status. Vulnerability can also arise from childhood, displacement, exploitation, dependency, or a credible safety risk.

Do not facilitate human trafficking, forced labor, sexual exploitation, abuse, or the recruitment, transport, advertising, control, or monetization of an exploited person.

Criticism, counterspeech, journalism, advocacy, and political disagreement are not automatically harassment or hate. Coul considers targeting, repetition, power imbalance, exposed private information, protected status, context, intent, reach, automation, and the likelihood and severity of harm—not whether Coul agrees with a viewpoint.

Do not threaten or incite violence, give operational assistance for violent wrongdoing, recruit for or materially support terrorism or violent extremism, celebrate graphic abuse to encourage or intimidate, facilitate weapon use for unlawful harm, or coordinate a dangerous act.

Do not encourage suicide, self-harm, dangerous eating-disorder behavior, or dangerous challenges, or give actionable instructions that increase those risks. Coul is not an emergency or crisis service; contact local emergency services when danger is imminent.

News, documentary, prevention, recovery, historical, counterspeech, research, or educational context may be permitted when the material is necessary, proportionate, clearly framed, and not celebratory, recruiting, threatening, or operational. Graphic material may still be restricted to protect viewers and affected people.

Do not use Coul for phishing, advance-fee or romance scams, fraudulent fundraising, pyramid or Ponzi schemes, fake giveaways, identity scams, malicious lead generation, fabricated testimonials or reviews, fake endorsements, false scarcity, deceptive pricing, or another material misrepresentation.

Do not buy, sell, generate, or coordinate fake followers, views, likes, comments, saves, shares, reviews, watch time, or other influence indicators. Do not operate fake-account networks, deceptive engagement pods, coordinated inauthentic behavior, or automation intended to manipulate a platform's ranking, ad, recommendation, or integrity systems.

Authentic calls to action, collaborations, contests, paid distribution, and community participation are allowed when truthful, consensual, properly disclosed, and permitted by the destination. AI-generated testimonials, avatars, or influence indicators must not be presented as genuine people or experience.

Do not create, target, schedule, or publish unlawful, unfair, deceptive, or unsubstantiated advertising; hide a sponsorship, affiliate relationship, material connection, risk, fee, eligibility term, or product limitation; fabricate an endorsement; or target unlawfully based on a protected or sensitive characteristic.

Do not promote illegal goods or unlawfully supplied weapons, drugs, controlled substances, tobacco or vaping products, gambling, adult services, counterfeit goods, financial products, health products, or another age-, location-, license-, or platform-restricted offering. Lawful restricted-goods content still requires every license, audience control, warning, disclosure, and destination permission.

Possess reasonable substantiation before making an objective claim. An AI edit must not crop, remove, obscure, contradict, or make unreadable a sponsorship, affiliate, risk, eligibility, safety, or product disclosure.

Lawful general education, news, personal experience, awareness, and commentary about health or finance are not prohibited merely by topic. Do not use Coul to impersonate a qualified professional, claim false credentials, or present generated content as individualized professional advice when authorization and appropriate review are required.

For health content, do not make false-cure, guaranteed-result, or unsubstantiated safety or efficacy claims; diagnose an identifiable person; unlawfully promote a prescription or controlled substance; or provide individualized treatment or prescribing instructions without required authority. A disclaimer does not cure a misleading overall impression.

For financial content, do not guarantee returns; manipulate a market; operate a pump-and-dump, fraudulent investment, or recovery scheme; use stolen or inside information; hide paid securities promotion or a conflict; make a deceptive risk claim; or provide unauthorized regulated advice, solicitation, or brokerage. Include required compensation, risk, conflict, and licensing disclosures.

Do not use a Coul viral score, recommendation, output, inferred trait, public-profile record, competitor analytic, engagement metric, or AI summary to make or materially support a decision about another person's employment, worker evaluation, housing, credit, lending, insurance, education, medical treatment, immigration, public benefit, law-enforcement treatment, legal service, or another significant entitlement, opportunity, or access to an essential service.

Coul's creator tools are not designed or validated for eligibility, risk, identity, truthfulness, criminality, health, emotion, or suitability determinations about a person. Human review does not make a prohibited high-impact use acceptable, and a public profile does not authorize that use.

Coul does not restrict lawful political opinion merely because of viewpoint. Do not use Coul for voter intimidation or suppression; knowingly false voting time, place, eligibility, or manner information intended to prevent participation; impersonation of an election official or candidate; fraudulent political fundraising; unlawful campaign-finance conduct; coordinated foreign interference; or undisclosed materially deceptive synthetic political media.

Do not use sensitive personal data for unlawful political profiling or microtargeting. Follow every sponsor, payer, targeting, transparency, recordkeeping, archive, synthetic-media, and platform requirement that applies to political or issue communications.

At launch, paid, sponsored, boosted, in-kind, or otherwise compensated advertising for or against a candidate, officeholder, political party, ballot measure, referendum, or election outcome—and automated election persuasion—are prohibited until Coul has implemented and verified jurisdiction, sponsor, disclosure, record, targeting-consent, age, provenance, and destination-platform controls. Do not route prohibited promotion through a creator, template, affiliate, workflow, or nominally organic post. Lawful uncompensated personal commentary, journalism, documentary work, and civic education remain allowed subject to the rest of this policy.

Do not create, deploy, distribute, or facilitate malware, ransomware, spyware, phishing kits, credential theft, exploit deployment, destructive code, unauthorized surveillance, service disruption, prompt injection intended to exfiltrate secrets, or evasion of security, policy, rate, access, or model-safety controls.

Do not scrape or collect private, login-gated, blocked, deleted, age-restricted, or otherwise unauthorized data; evade CAPTCHAs or rate limits; stuff credentials; enrich or broker personal data unlawfully; re-identify protected data; reverse engineer where prohibited; or resell Coul or provider access. Coul's approved in-product competitor feed is not permission to independently scrape a source.

Proportionate defensive security testing, education, and research may be permitted only with authorization, scoped targets, appropriate safeguards, no harmful deployment, and responsible handling of findings. Do not use a research label to target a system or person without permission.

Do not use nodes, templates, schedules, connected accounts, or agents for bulk spam, repetitive deceptive posts, hashtag abuse, fake-account networks, influence manipulation, policy-shopping between destinations, provider-limit evasion, or rapid publication of unreviewed risky content.

You remain responsible for every scheduled or automatically produced post. Review the target account, audience, time zone, claims, evidence, rights, disclosures, accessibility, provenance, and destination labels. Pause automation promptly when permissions, content, a campaign, platform rules, or safety conditions change.

Before launch, Coul must apply appropriate policy checks before public-template approval and before direct or automated publishing, re-check materially changed output and destination settings, and prevent workspace-role bypass of final approval. A provider content-policy or spam rejection must be terminal: it must not be automatically retried, repackaged, published manually, or routed to another destination to evade the block. Publication can resume only after a compliant edit or a legitimate provider appeal resolves the issue.

Coul may permit otherwise sensitive material for bona fide journalism, documentary work, education, prevention, recovery, counterspeech, historical context, public-interest research, satire, fiction, art, or authorized security testing when the use is lawful, necessary to the context, proportionate, appropriately labeled, and does not provide operational assistance or evade safeguards.

Context is assessed from the request, media, text, audience, intent, likely effect, distribution, and safeguards—not from a label alone. Coul may limit visibility, generation detail, automation, or publication even when retaining content for a legitimate record or report.

No contextual exception permits child sexual exploitation, grooming, non-consensual intimate imagery, credible threats, trafficking, extortion, malware deployment, or another act Coul is legally required to prohibit or report.

Users and non-users must be able to report a template, public output, account, impersonation, privacy exposure, infringement, scam, threat, child-safety concern, or non-consensual intimate image without creating a Coul account where law requires it. The report should identify the source URL, Coul item or job, account, reason, and only the minimum evidence needed.

Before launch, Coul must add monitored urgent categories for child exploitation, imminent threats, non-consensual intimate imagery, trafficking, malware, impersonation, and election interference, as well as ordinary categories and separate copyright and privacy routes. Current generic template reasons are not enough.

Coul may temporarily quarantine a public template or pause a job while reviewing a credible report. One report or an automated/provider signal can justify containment but is not by itself a final finding. The workflow must resist abusive reports, protect reporter identity where appropriate, issue a case identifier and status, minimize retained evidence, and support restoration after error.

Depending on severity and risk, Coul may refuse a request; warn a user; reduce visibility; quarantine or remove a template; stop a queued job; disable automation, an integration, connected account, feature, workspace, or user; limit rate or access; preserve evidence; suspend or terminate an account; and make a legally required report.

Coul considers intent, severity, recurrence, reach, automation, vulnerability of affected people, prior notice, remediation, and risk of continued harm. Immediate action can be appropriate for imminent threats, child exploitation, non-consensual intimate imagery, malware, fraud, evidence destruction, or legal obligation; otherwise Coul should provide notice and a reasonable cure where safe and appropriate.

Except where law or safety prevents disclosure, an enforcement notice should identify the rule, relevant facts, action scope and duration, whether automation materially contributed, available preservation or export rights, and appeal route. Appeals should be free and accessible and receive a fresh evaluation of the evidence and policy by a sufficiently independent human reviewer with authority to reverse. That reviewer must not merely rubber-stamp or rely solely on the original human decision or automated signal and must be able to restore content or access.

Coul does not promise that every upload, prompt, template, schedule, or post is reviewed or that every violation will be detected. Provider safety blocks, malware scans, user reports, platform rejections, and automated classifiers are signals—not substitutes for appropriate product controls and human judgment.

Before launch, Coul must activate general safety, copyright, privacy, non-consensual intimate imagery, child-safety, emergency, and appeal channels; create tracked cases, reason codes, notices, reviewer notes, reversals, restoration, and transparent retention; and train reviewers on urgent harm, IP, context, medical and financial claims, and elections.

Coul must configure and test AI safety settings across generation and editing; enable malware scanning fail-closed where appropriate without calling it content moderation; add pre-publish and template checks; verify workspace permissions and terminal provider-policy failures; implement intimate-image and legally required child-safety escalation; enforce the temporary political-ad restriction; preserve ad disclosures; support synthetic-media provenance; and verify destination and provider rules.

Coul must also verify its legal entity, geographic address, monitored contacts, law-enforcement process, statutory reporting obligations, regional hosting or online-platform duties, and policy-version records. Questions may be sent to legal@coul.app only after the mailbox is verified and monitored. Never email prohibited imagery, passwords, access tokens, or unnecessary identity documents.