Community Covenant

On this page

PREAMBLE: THE DOCTRINE OF SOVEREIGNTY

ARTICLE I: VIOLENCE AND CRIMINAL BEHAVIOR

ARTICLE II: SAFETY AND HUMAN DIGNITY

ARTICLE III: OBJECTIONABLE CONTENT & REALITYS

ARTICLE IV: NETWORK INTEGRITY & AUTHENTICITY

ARTICLE V: INTELLECTUAL PROPERTY

ARTICLE VI: ENFORCEMENT & THE SHADOW PROTOCOL

ARTICLE VII: APPEALS & OVERSIGHT

ARTICLE VIII: AMENDMENTS & EVOLUTION

PREAMBLE: THE DOCTRINE OF SOVEREIGNTY

1. The Nature of The Network

Kemet (hereinafter "The Network") is a decentralized communication utility and social media protocol built upon three foundational principles: cryptographic sovereignty, immutable identity, and non-custodial data ownership. Unlike legacy platforms that function as publishers, curators, or editors of user content—exercising editorial discretion over what may or may not be said—Kemet functions fundamentally as a neutral carrier of encrypted information, similar to the postal service or telephone network in traditional communications.

However, neutrality does not imply anarchy. The structural integrity of The Network—and the safety of the Nodes (user accounts) within it—requires certain minimum standards of conduct. This Community Covenant (the "Covenant") establishes those standards, distinguishing between the immutable mathematical laws enforced by cryptography and the community standards enforced by consensus and administrative action.

2. The Distinction: Code vs. Policy

This Covenant draws a fundamental distinction between two categories of rules:

A. Immutable Law (Enforced by Cryptography) These are technical constraints built into the Kemet Secure Messaging Protocol (KSMP) and the architecture of The Network itself: • We cannot read your Direct Messages. The encryption is end-to-end; only you and your intended recipients possess the keys. • We cannot delete your Vault Key or recovery phrase. These exist only on your device. • We cannot prevent you from creating a Node or sending encrypted messages. The mathematics of public-key cryptography makes this impossible. • We cannot recover your account if you lose your keys. This is not policy; it is mathematical reality.

These limitations are features, not bugs. They ensure your cryptographic sovereignty.

B. Community Law (Enforced by Consensus and Administration) These are the standards set forth in this Covenant. While we cannot prevent you from existing or communicating privately, we can and will restrict the propagation of your content through the Public Feed, Group Discovery, and Global Search if you violate the terms herein. We can terminate your access to public features of The Network. We can preserve metadata for law enforcement handoff where credible threats to life exist.

This Covenant governs Community Law. It applies to all users of The Network, regardless of jurisdiction, nationality, or status. By using The Network, you accept and agree to be bound by this Covenant.

3. The "Common Sense" Doctrine

Legacy platforms rely on exhaustive, bureaucratic lists of prohibited words, attempting to regulate speech through mechanical keyword filtering. This approach fails because: • Context matters. The word "bomb" in "this movie is the bomb" differs fundamentally from "I planted a bomb." • Language evolves. Slurs, code-words, and euphemisms constantly shift. • Technical circumvention is trivial. Users substitute characters, use leetspeak, or employ euphemisms.

Kemet rejects this approach. We employ the Common Sense Doctrine: we adjudicate behavior based on Intent and Impact, not mere vocabulary. We examine: • What was the apparent purpose of the content? • What was the likely effect on recipients and The Network? • What was the context (historical, educational, satirical, or malicious)?

Attempting to bypass these rules through technicalities ("Wikilawyering")—such as claiming "I didn't use the exact prohibited word" or "this is technically legal in my jurisdiction"—will be treated as an act of bad faith and may result in immediate Node Termination. Kemet Protocol Governance retains sole and absolute discretion in interpreting user intent, context, and determining what constitutes a violation of this Covenant.

4. Scope and Application

This Covenant applies to: • All Public Content (posts, comments, media, links) • All Group communications in public or discoverable Groups • All usernames, display names, avatars, and profile information • All use of The Network's discovery, search, and recommendation features • All attempts to circumvent enforcement or manipulate The Network

This Covenant does NOT apply to: • End-to-end encrypted Direct Messages between consenting parties • Private, encrypted Group communications in undiscoverable, invite-only Groups • Content stored solely on your device and never transmitted

However, note that while we cannot access encrypted private communications, we can and will act if: • A recipient reports content to us (providing decryption keys for that specific content) • Law enforcement provides valid legal process compelling a recipient to disclose • Metadata analysis reveals patterns of criminal coordination (e.g., bulk recruitment for terrorism)

5. Severability and Interpretation

If any provision of this Covenant is held invalid or unenforceable, the remaining provisions shall continue in full force and effect. This Covenant shall be interpreted to preserve the safety and integrity of The Network while maximizing individual sovereignty consistent with that safety.

ARTICLE I: VIOLENCE AND CRIMINAL BEHAVIOR

The Network tolerates offensive speech. It does not tolerate entropy, destruction, or the coordination of physical harm.

Section 1.1: Terrorism and Violent Extremism

Kemet maintains a Zero Tolerance policy regarding terrorism and violent extremism.

A. Definition For purposes of this Covenant, "Terrorism" means: • Acts intended to cause death or serious bodily injury to civilians • Acts intended to destabilize or destroy fundamental political, constitutional, economic, or social structures • Acts intended to intimidate a population or compel a government or international organization to act or abstain from acting

We define terrorist organizations based on the consolidated lists of: • The United Nations Security Council • The U.S. Department of State Foreign Terrorist Organizations list • The European Union's list of terrorist organizations • Comparable lists maintained by major international bodies

B. Prohibited Acts The following acts constitute violations of this Covenant and will result in immediate Termination, law enforcement handoff, and permanent blacklisting:

1. Recruitment: Using The Network to solicit, recruit, or facilitate membership in designated terrorist organizations. This includes: • Posting recruitment materials or propaganda • Direct messaging users with recruitment appeals • Creating Groups or channels for terrorist organization coordination • Providing contact information for terrorist recruiters

2. Material Support: Providing financial, logistical, or technical assistance to terrorist organizations, including: • Fundraising or soliciting donations • Sharing operational security guides • Providing technical infrastructure assistance • Facilitating travel or procurement

3. Glorification: Creating, sharing, or amplifying content that praises, celebrates, or encourages specific acts of terrorism, including: • Celebrating bombings, mass shootings, beheadings, or kidnappings • Memorializing terrorists as "martyrs" or "heroes" • Creating artistic content that aestheticizes terrorist violence • Sharing execution videos or gore for propagandistic purposes

4. Instruction: Disseminating blueprints, guides, tutorials, or instructional materials on: • Manufacturing explosives, improvised explosive devices (IEDs), or incendiary devices • Production of chemical, biological, or radiological weapons • Construction of illicit firearms ("ghost guns") or conversion of legal firearms to automatic • Evasion of security screening or law enforcement detection • Operational security for criminal or terrorist activities

C. Exceptions The following activities do NOT violate this section: • Academic research on terrorism (published in educational context) • News reporting on terrorist activities (factual, not celebratory) • Counter-extremism or deradicalization efforts • Artistic works that condemn or analyze terrorism (not glorify) • Discussion of political causes that may be associated with terrorism (without supporting terrorist methods)

D. Enforcement Violations of this section result in: • Immediate Node Termination (no warning) • Permanent blacklisting of Public Key, Device Fingerprint, and IP Hash • Preservation of all available metadata (IP hashes, timestamps, device information) for law enforcement handoff • Potential reporting to relevant authorities where credible threats to life exist

Section 1.2: Organized Hate and Violent Groups

A. Scope We do not allow groups or individuals that exist primarily to coordinate violence, intimidation, or discrimination based on protected characteristics (race, ethnicity, national origin, religion, sexual orientation, gender identity, disability).

This includes but is not limited to: • Neo-Nazi organizations and cells • White supremacist networks and accelerationist groups • Violent separatist militias or paramilitary organizations • Organized criminal gangs or syndicates • Violent extremist movements (regardless of political ideology)

B. Prohibited Content and Conduct 1. Organizational Activity: Creating or maintaining Groups, channels, or networks for: • Coordinating violent activities • Sharing operational plans or target lists • Recruiting for violent group membership • Fundraising for violent operations

2. Hate Symbols: The display of hate symbols is prohibited unless used in a clear, verifiable Historical, Educational, or Documentary context (HED exception): • Swastikas (except in Buddhist/Hindu religious context, or historical documentation) • SS Runes, Totenkopfs, or other Nazi iconography • KKK insignia or burning cross imagery • Terrorist organization flags or emblems (ISIS, Al-Qaeda, etc.)

Note: Using hate symbols "ironically," "for shock value," or "as a joke" is not a valid defense. If the symbol appears without clear HED context, it constitutes a violation.

3. Targeted Threats: Threats of violence directed at: • Individuals based on protected characteristics • Religious institutions or community centers • LGBTQ+ spaces or events • Racial or ethnic communities

C. The HED Exception (Historical, Educational, Documentary) Content featuring hate symbols or extremist imagery may be permitted if: • It is clearly labeled as historical documentation, educational material, or documentary journalism • It does not celebrate or promote the ideology represented • It provides context that condemns or analyzes the ideology • It is not shared in spaces dedicated to promoting that ideology

Examples: ✓ A history teacher sharing a documentary about the Holocaust with educational context ✓ A journalist reporting on white supremacist movements with analysis ✗ A user posting Nazi imagery with "ironic" captions in a meme group ✗ A channel sharing ISIS propaganda "to understand the enemy"

D. Distinction from Offensive Speech Mere offense is not a violation. The following are generally permitted: • Offensive opinions or political views (even if widely considered bigoted) • Debate about immigration policy, religious doctrine, or social issues • Criticism of ideologies, movements, or belief systems • Satire or parody of hate groups (if clear and not promoting)

The line is drawn at: • Coordination of violence • Targeted harassment campaigns • Use of hate symbols without HED context • Explicit calls for discrimination or exclusion

Section 1.3: Promoting Criminal Activities

The Network is not a marketplace, coordination platform, or how-to guide for illegal goods and services.

A. Prohibited Commerce The following commercial activities are strictly prohibited:

1. Narcotics and Controlled Substances: • Sale, purchase, or distribution of illegal drugs • Facilitation of drug transactions (connecting buyers and sellers) • Sharing information on procurement or manufacturing of illegal drugs • Promotion of drug use (except in recovery or harm reduction contexts)

Note: Discussion of drug policy, legalization debates, or harm reduction is permitted. The prohibition applies to transactional or facilitating behavior.

2. Weapons and Explosives: • Private sale of firearms, ammunition, or explosives • Sale of firearm parts or conversion kits • Instructions for manufacturing weapons • Coordination of straw purchases or evasion of background checks

3. Human Trafficking and Exploitation: • Any facilitation of sexual exploitation or forced labor • Recruitment for prostitution or escort services • Sharing contact information for traffickers • Arranging travel for exploitative purposes

4. Cybercrime and Fraud: • Distribution of malware, ransomware, or spyware • Sale of stolen credentials ("dox logs," "combolists") • Phishing kits or social engineering guides • Credit card fraud, identity theft guides • Sale of compromised accounts or "cracked" software

5. Financial Crimes: • Money laundering services • Unlicensed money transmission • Ponzi schemes or investment fraud • Tax evasion schemes

B. Permitted Discussion The following are generally permitted (context matters): • Academic discussion of criminal justice, criminology, or deviance • Cybersecurity research and vulnerability disclosure • Historical analysis of criminal organizations • Legal advice regarding criminal charges • Whistleblowing about criminal activity (with legal protections)

C. Jurisdictional Variations Laws vary by jurisdiction. What is legal in one country may be illegal in another. We enforce based on: • The jurisdiction where The Network operates (Delaware law) • International consensus (e.g., UN conventions on drugs, weapons, human rights) • The most restrictive applicable standard where The Network has presence

Do not assume that because something is legal in your location, it is permitted on The Network.

ARTICLE II: SAFETY AND HUMAN DIGNITY

Kemet prizes sovereignty, but sovereignty cannot exist where human dignity is violated through exploitation or where vulnerable individuals are driven to self-destruction. This Article establishes the non-negotiable boundaries of safety.

Section 2.1: Child Sexual Exploitation (CSAM)

THIS IS THE ABSOLUTE RED LINE. Kemet has Zero Tolerance for any content or activity involving the sexual exploitation of children.

A. Definition "Child Sexual Abuse Material" (CSAM) means any visual depiction of sexually explicit conduct involving a minor (under 18 years of age). This includes: • Photographs, videos, or digital images • Computer-generated imagery (CGI) that is indistinguishable from real children • Drawings, cartoons, or animations depicting sexual acts with minors ("loli," "shota") • Modified or "morphed" images that sexualize minors

B. Prohibited Acts The following constitute immediate, irreversible Termination and mandatory law enforcement reporting:

1. Possession, Distribution, or Production: Any sharing, storing, or creation of CSAM.

2. Solicitation and Grooming: Attempting to establish sexual contact with minors, including: • Sexualized communication with minors • Attempting to procure sexual images from minors • Arranging meetings for sexual purposes • "Grooming" behaviors (building trust for sexual exploitation)

3. Normalization: Content that attempts to normalize adult-minor sexual relationships, including: • Pro-pedophile advocacy • "Minor Attracted Person" (MAP) community promotion • Sharing "research" or literature supporting child sexual abuse

C. Detection and Enforcement • We utilize client-side and transit-layer perceptual hashing (PhotoDNA or equivalent proprietary technology) to detect known CSAM against NCMEC (National Center for Missing and Exploited Children) databases. • Matches are automatically reported to NCMEC CyberTipline, including IP Hash, Timestamp, and Device ID. • Data is moved to forensic cold-storage partition for law enforcement access. • Node is immediately and irreversibly Terminated. • Device fingerprint is globally blacklisted.

D. No Exceptions There are no exceptions to this policy. Claims of "artistic expression," "historical documentation," or "research" will not be accepted. If you encounter CSAM, report it immediately. Do not attempt to "investigate" or "collect evidence"—this constitutes possession.

Section 2.2: Suicide and Self-Injury

We do not allow content that promotes, glorifies, or provides detailed instructions for self-destruction.

A. Prohibited Content 1. Detailed Instructions: Step-by-step guides to suicide methods, including: • Specific dosages for overdose • Ligature techniques • Methods of self-harm with high lethality • "Suicide kits" or procurement guides

2. Glorification and Aestheticization: Content that presents suicide or self-harm as: • Desirable or romantic • A solution to problems • An expression of identity or belonging • Aesthetic or artistic (without clear anti-suicide message)

3. Real-Time Depictions: Images or videos of: • Active self-harm (cutting, burning, overdose) • Suicide attempts in progress • Immediate aftermath of suicide (before authorities/family notification)

4. Pro-ED Content: "Pro-Ana" (anorexia) or "Pro-Mia" (bulimia) content that: • Encourages eating disorders as a lifestyle • Shares "thinspiration" or body-checking imagery • Provides tips for hiding disorders or avoiding treatment • Celebrates dangerous weight loss

B. Permitted Content The following are generally permitted and encouraged: • Personal stories of recovery and survival • Mental health advocacy and resource sharing • Crisis hotline information • Discussion of therapy, medication, or treatment • Artistic expression that clearly condemns suicide (not glorifies) • Academic or clinical discussion of self-harm (educational context)

C. Intervention and Safety Resources • Users searching for terms related to self-harm will be served a Safety Interstitial directing them to local crisis resources before results are displayed. • We partner with crisis organizations to provide in-app support. • Users can report concerning content for welfare checks (metadata preserved for emergency services if imminent risk identified).

Section 2.3: Adult Nudity and Sexual Activity

Kemet is not a pornography platform. While we respect bodily autonomy and sexual expression, public spaces on The Network are not appropriate for explicit sexual content.

A. Prohibited Content 1. Hardcore Pornography: • Visible genitalia with sexual intent • Sexual acts (penetration, oral sex, masturbation) • Ejaculation or sexual fluids depicted sexually • Sex toys used in explicit sexual context

2. Non-Consensual Intimate Imagery (NCII): • "Revenge porn" or intimate images shared without consent • Hidden camera or "upskirt" photography • Deepfake pornography of real persons (without consent) • Voyeuristic content

Note: NCII is a criminal violation in many jurisdictions. We will report to law enforcement and preserve evidence.

3. Sexual Violence: • Depictions of sexual assault or rape (real or simulated, unless clear educational/documentary context) • Content that eroticizes non-consent • Bestiality or sexual acts with animals

B. Permitted Content (With Appropriate Tagging) The following may be permitted if properly tagged and contextually appropriate:

1. Artistic Nudity: • Fine art photography or sculpture depicting nudity • Classical paintings or drawings • Performance art with nudity (non-sexual context)

2. Educational Content: • Medical or anatomical diagrams • Sex education materials (clinical, not pornographic) • Breastfeeding education

3. Protest and Expression: • Nudity as political protest (e.g., FEMEN-style protests) • Body positivity movements (non-sexualized) • Breastfeeding in public (normalized, not fetishized)

C. The NSFW Tag Requirement Users posting borderline content (suggestive but not explicit; artistic nudity; medical content) MUST apply the [NSFW] (Not Safe For Work) tag.

Failure to tag results in: • Content being locked (hidden from feeds) • Trust Score reduction • Potential Suspension for repeated violations

D. Private vs. Public Distinction • Private, encrypted communications between consenting adults: Not moderated (we cannot access) • Public posts, discoverable Groups, profiles: Subject to this section • If you wish to share sexual content, use private, encrypted channels with trusted recipients

Section 2.4: Bullying and Harassment

We do not police "mean words" or isolated insults. We police Targeted Harassment Campaigns—coordinated, sustained attacks designed to drive individuals from The Network or cause severe distress.

A. Prohibited Conduct 1. Doxxing: Broadcasting private, non-public information about an individual with intent to enable harassment, including: • Home addresses • Personal phone numbers • Private email addresses (not publicly associated with the individual) • Workplace information (if not publicly disclosed by the individual) • Family member information • Financial information (bank accounts, credit cards)

Note: "Doxxing" requires intent to harass. Publishing publicly available information (e.g., a public business address) is not doxxing.

2. Swatting: Falsely reporting emergencies to send armed police or emergency services to a victim's location. This is a felony in many jurisdictions and will result in immediate Termination and law enforcement handoff.

3. Dogpiling: Coordinating a mass attack on a specific user, including: • Organizing Groups or hashtags to target an individual • Encouraging followers to report an account en masse (brigading) • Coordinated insult campaigns designed to overwhelm • "Cancel culture" campaigns that cross into harassment

4. Stalking: Persistent, unwanted attention that causes fear or distress, including: • Following a user across multiple posts to insult or threaten • Creating multiple accounts to circumvent blocks • Monitoring and documenting a user's activity obsessively • Sending unwanted messages after being told to stop

B. The "Block" Sovereign Right Users have the absolute right to block any Node. Blocking means: • The blocked user cannot see your content • The blocked user cannot message you • The blocked user cannot interact with your posts

Circumventing a block by creating new Nodes ("Ban Evasion") constitutes Harassment and will result in device-level banning.

C. Permitted Speech The following are generally NOT harassment: • Isolated insults or arguments • Criticism of public figures or public statements • Disagreement with ideas or opinions • Satire or parody of public personas • Reporting genuine violations of this Covenant

The line is drawn at: • Coordination (multiple users acting together) • Persistence (continuing after being blocked or told to stop) • Disclosure of private information • Intent to cause fear or drive from platform

ARTICLE III: OBJECTIONABLE CONTENT & REALITY

Kemet differentiates between "Graphic Violence" and "Graphic Reality." We refuse to sanitize the world for user comfort, but we will not host sadistic entertainment.

Section 3.1: Graphic Violence & War Footage (The "Reality Clause")

Unlike legacy platforms that sanitize war to make it "advertiser-friendly," Kemet permits the documentation of conflict, violence, and human rights abuses. We believe in the right to know and the duty to bear witness.

A. Permitted Content The following is permitted, with mandatory tagging:

1. Combat Footage: • Drone strikes, firefights, military operations (from news or documentary sources) • Body camera footage from military or police • Aftermath of attacks (destruction, casualties)

2. Evidence of Human Rights Abuses: • Police brutality documentation • War crimes evidence • Prison abuse footage • Forced displacement documentation

3. Medical and Educational: • Surgical procedures (if educational) • Forensic documentation • Accident reconstruction

B. Strict Conditions 1. Mandatory Tagging: All such content MUST be flagged [GRAPHIC] or [WAR]. Failure to tag results in Shadow Containment.

2. No Sadism: Content that depicts torture, mutilation, or killing solely for shock value or sadistic pleasure is prohibited. The presence of suffering must serve documentary, evidentiary, or educational purpose—not entertainment.

3. No Celebration: Captions or commentary that mock victims, celebrate their suffering, or dehumanize them transform the content from "Documentary" to "Hate Speech," which is prohibited.

Examples: ✓ "Documenting Russian airstrike on civilian shelter, Kharkiv 2024" [WAR] ✗ "LOL look at these guys getting rekt 💀" [VIOLATION]

4. Default View: Such content is blurred by default. Users must opt-in to view it via a "Show Content" click. This protects accidental exposure while preserving access.

C. Context Matters The same image can be documentary or glorification depending on context: • A photo of a war casualty with caption "Never forget the cost of war" = Documentary • The same photo with caption "This is what happens to traitors" = Glorification/Violation

Section 3.2: Hate Speech (The Strict Definition)

We do not use the vague, expansive "Hate Speech" definitions prevalent on legacy platforms, which often conflate offense with harm. We use a strictly defined Dehumanization Standard.

A. Prohibited: Dehumanization Speech that dehumanizes individuals or groups based on race, ethnicity, religion, disability, or gender identity by comparing them to:

1. Insects, animals, or pests: • "Rats," "cockroaches," "vermin," "infestation" • "Swarm," "invasion," "plague"

2. Filth, bacteria, or viruses: • "Parasites," "disease," "cancer," "tumor" • "Cleanse," "purge," "disinfect"

3. Sub-humans or objects: • "Subhuman," "mongrels," "mongoloids" • "Property," "tools," "livestock" • "Replaceable," "disposable"

B. Prohibited: Calls for Violence • "[Group] should be exterminated" • "[Group] must be removed from society" • "[Group] should be segregated" • "Death to [group]" • "[Group] has no right to exist"

C. Permitted: Offensive Speech The following are generally permitted, even if widely considered offensive: • Insults and mockery ("You're an idiot," "Your religion is stupid") • Blasphemy and criticism of religion • Offensive opinions about political issues • "I hate [group]" (expression of emotion, not call for action) • Stereotypes and generalizations (offensive but not dehumanizing)

D. The Kemet Stance You have the right to be offensive. You do not have the right to incite genocide.

We protect the expression of hatred (emotion) but prohibit the organization of harm (action). The line is drawn at dehumanization and incitement, not at offense.

ARTICLE IV: NETWORK INTEGRITY & AUTHENTICITY

Kemet functions on the Proof of Work of human attention and authentic engagement. We prohibit the use of automation and deception to distort the reality of the public square.

Section 4.1: Spam and Platform Manipulation

We define "Spam" not merely as unwanted content, but as Artificial Amplification—attempts to manipulate the perceived popularity or consensus of content through inauthentic means.

A. Prohibited Acts

1. Synthetic Engagement ("Engagement Farming"): • Buying, selling, or trading likes, reposts, followers, or comments • Using botnets or click farms to inflate engagement metrics • Coordinating "like parties" or "follow trains" for artificial growth • Automated liking/reposting without human decision

2. Coordination and Astroturfing: • Using multiple Nodes (sockpuppets) to artificially amplify a specific hashtag or topic • Creating fake "grassroots" movements (astroturfing) • Coordinating mass posting of identical or similar content • Organized brigading of comments or reviews

3. Scripted Interaction: • Automating actions (auto-liking, auto-replying, auto-following) without using the official, rate-limited API • Using scripts to evade rate limits • Automated account creation or management

4. Scams and Fraud: • Financial scams (advance fee fraud, Ponzi schemes, fake investment opportunities) • Romance scams or catfishing for financial gain • Impersonation of customer support to steal credentials • Fake giveaways or lottery scams

B. Detection Methods Our Heuristic Engine analyzes: • Posting velocity (humanly impossible rates, e.g., 100 posts/minute) • IP hash rotation (rapid switching to evade limits) • Graph clustering (networks of accounts acting in concert) • Behavioral patterns (perfect timing, repetitive text, lack of variation) • Device fingerprint anomalies (emulation, virtualization)

C. Consequences • Automatic Shadow Containment for suspected bot/Script activity • The Node remains active to the user but signals are dropped by network peers • Duration: Until Trust Score recovers (via positive human interactions) or permanent for confirmed bots • No notification to prevent bot adaptation

The Network employs KSLP (Kemet Secure Link Protocol) to analyze, rate, and shorten external URLs posted in public content. Malicious or unsafe links are flagged and users are warned before redirection. KMSP (Kemet Media Safety Protocol) performs perceptual hashing on public images and videos to detect and block abusive content. These systems operate on public content only and cannot access encrypted private communications.

Section 4.2: Impersonation and Deception

Identity on Kemet is sovereign, but it must not be fraudulent.

A. Prohibited Impersonation

1. Misappropriation of Identity: • Using the name, image, or likeness of another individual to mislead others into believing you are that person • Creating fake profiles of public figures, celebrities, or executives • Impersonating journalists, government officials, or corporate representatives • "Catfishing"—using photos of others to create romantic or social deception

2. Brand Phishing: • Creating Nodes that mimic official Kemet communications or administrative accounts • Fake "Kemet Support" or "Kemet Security" accounts • Impersonating banks, crypto exchanges, or trusted services to steal Vault Keys

3. Deepfakes and Synthetic Media: • Posting AI-generated audio/video of real persons without clear [AI] or [PARODY] label • Voice cloning for fraud or deception • Synthetic pornography of real persons (also violates Section 2.3)

B. Permitted Parody and Satire The following are allowed if clearly labeled: • Parody accounts with "(Parody)" or "(Satire)" in the bio • Caricature or exaggeration for comedic effect • Commentary accounts (e.g., "Elon Musk (Not)" with clear indication) • Fan accounts that do not claim to be the official person

C. Pseudonymity is Protected You are free to use a handle that is not your legal name. You are free to be anonymous. We do not require "real name" policies.

The prohibition is on impersonation (claiming to be someone you're not), not on anonymity (choosing not to reveal who you are).

Section 4.3: Misinformation and Civic Integrity

Unlike legacy platforms, Kemet does not act as an "Arbiter of Truth." We do not fact-check opinions. We do not censor conspiracy theories. We do not intervene in scientific debates.

However, we protect the Structure of civic discourse by prohibiting specific harms:

A. Prohibited: Census and Voting Interference • Distributing false information about how, when, or where to vote in official elections • "Text your vote to this number" scams • False claims about voting eligibility or requirements • Intimidation at polling places (coordinated through platform)

B. Prohibited: Public Health Hoaxes (Imminent Harm) • Promoting cures or treatments that pose immediate, scientifically proven risk of death or severe injury • "Drink bleach to cure COVID" • "Vaccines contain microchips" (if accompanied by calls for violence against vaccinated) • Instructions for DIY medical procedures that cause harm

Note: We permit heterodox science, conspiracy theories, and political dissent about public health. We only intervene when speech creates a verifiable, kinetic risk of severe injury or death.

C. Permitted: Speech We Do Not Censor • Conspiracy theories (flat earth, moon landing denial, etc.) • Political dissent and protest organizing • Scientific heterodoxy (challenging consensus) • Religious or philosophical claims • Satire and parody of news events • Opinion, commentary, and analysis

The Kemet Stance: We protect the right to be wrong. We only intervene when being wrong causes imminent, kinetic harm.

ARTICLE V: INTELLECTUAL PROPERTY (DMCA)

Kemet respects the creative sovereignty of authors, artists, and innovators. We comply with the Digital Millennium Copyright Act (DMCA) and equivalent international copyright frameworks.

Section 5.1: The Notice and Takedown Procedure

A. Reporting Infringement Copyright holders or authorized agents may submit a DMCA Takedown Notice to legal@kemet.network. The notice must contain:

1. Physical or electronic signature of the copyright owner or authorized agent 2. Identification of the copyrighted work claimed to be infringed (or a representative list if multiple works) 3. Identification of the infringing material and its location on The Network (exact URL/Post ID) 4. Contact information (address, telephone number, email) 5. Good faith belief statement that use is not authorized by copyright owner, agent, or law 6. Accuracy statement under penalty of perjury 7. Confirmation that the notifying party is authorized to act on behalf of the copyright owner

B. Action Upon Receipt Upon receipt of a valid notice: • We will promptly disable access to the specific content in the Public Feed • We will notify the Node operator of the takedown • We will provide information on filing a Counter-Notice • We will preserve the takedown notice for legal purposes

C. Counter-Notice Procedure The Node operator may file a Counter-Notice if they believe the takedown was erroneous or the content is non-infringing (e.g., Fair Use). The Counter-Notice must include: • Physical or electronic signature • Identification of removed content and its prior location • Good faith belief statement that removal was mistake or misidentification • Contact information and consent to jurisdiction of federal court • Confirmation of willingness to accept service of process

If a valid Counter-Notice is filed: • We will notify the original claimant • Content will be restored in 10-14 business days unless the claimant files a court order • If court order filed, content remains down pending legal resolution

Section 5.2: The Repeat Infringer Policy

A. Three Strikes Nodes that receive three (3) valid copyright strikes within a rolling 6-month period will be subject to Termination.

B. Scope This policy applies to Public Feed and Public Groups. Private, end-to-end encrypted communications cannot be scanned for copyright infringement, and Kemet cannot adjudicate the contents of sealed messages.

C. Restoration Strikes expire after 6 months. Nodes with clean records for 6 months may petition for early restoration of full privileges.

Section 5.3: Fair Use and Exceptions

We acknowledge that not all uses of copyrighted material are infringing. Factors we consider in disputes: • Purpose and character of use (commercial vs. educational, transformative nature) • Nature of copyrighted work (factual vs. creative) • Amount and substantiality of portion used • Effect on potential market for original work

Parody, commentary, criticism, news reporting, teaching, scholarship, and research may qualify as Fair Use.

ARTICLE VI: ENFORCEMENT & VISIBILITY RESTRICTION

We do not use "Human Moderators" to read every post. We use verifiable algorithms and reputation math. This Article details the technical enforcement mechanisms.

Section 6.1: The Trust Score (The Algorithm)

Every Node possesses a dynamic Trust Score calculated by our Heuristic Engine. To ensure absolute neutrality and prevent platform bias, the weights, logic, and mechanics of the Trust Score algorithm are published as open-source documentation. While your specific numerical score remains hidden to prevent gamification by botnets, the mathematical rules governing the platform are fully transparent.

A. Calculation Factors The Trust Score is derived from:

1. Account Age: Older nodes receive higher baseline trust (reduces Sybil attacks) 2. Graph Density: Nodes followed by, or following, high-trust nodes gain trust (social proof) 3. Report Volume: Nodes receiving frequent blocks/reports from other users lose trust 4. Behavioral Entropy: Human behavior is messy, irregular, and inefficient. Perfect timing, repetitive text patterns, and mechanical consistency indicate automation and reduce trust 5. Content Quality Signals: Engagement with legitimate, high-quality content increases trust; association with spam, scams, or abuse reduces trust 6. Verification Status: Kemet+/Kemet++ subscribers receive trust bonuses (economic commitment signals legitimacy) 7. Device Fingerprint Consistency: Rapid device changes or emulator use reduces trust temporarily 8. Rate Limit Compliance: Respecting rate limits maintains trust; evasion attempts reduce trust

B. Score Range and Thresholds • 90-100: High Trust (full visibility, all features) • 70-89: Normal Trust (standard operation) • 50-69: Reduced Trust (rate limits, reduced algorithmic distribution) • 30-49: Low Trust (shadowing likely, manual review triggered) • 20-29: Critical Trust (Shadow Containment activated) • 0-19: Termination Threshold (human review for ban)

Section 6.2: Visibility Restriction (The Trust Quarantine)

When a Node's Trust Score drops below the critical threshold (20, the "Red Line"), the Visibility Restriction Protocol activates automatically.

A. Effect of Visibility Restriction • The user can still post, message, and use The Network normally • Direct Messages and private Group communications remain unaffected • Public posts are de-indexed from Global Search and Hashtags • Public posts do not appear in the algorithmic Public Feed • The Node is temporarily quarantined to protect the network from suspected automation or abuse

B. Purpose Visibility Restriction serves two functions: Network Protection: Prevents suspected botnets and spammers from amplifying content before human review. Rehabilitation: Human users who triggered restrictions inadvertently can recover through positive behavior (genuine engagement, no reports, time passing).

C. Duration Visibility Restriction lasts until: • The Trust Score recovers above 30 through positive interactions, OR • The Node is terminated for confirmed abuse, OR • Permanent quarantine is applied for confirmed automated botnets.

D. Notification Policy To comply with international digital rights standards and respect user autonomy, human users are notified of visibility restrictions. If your account is restricted, you will receive an automated system notice explaining that your Trust Score has dropped below the operational threshold, resulting in your content being de-amplified from public discovery. You will be provided with instructions on how to appeal or restore your Trust Score. (Note: Accounts algorithmically confirmed with 99% certainty to be automated spam networks may be restricted silently to prevent adversarial adaptation).

Section 6.3: Suspension and Termination (The Hard Ban)

For violations of Article I (Violence) or Article II (Safety), we escalate to hard enforcement.

A. Suspension Temporary "Read-Only" mode where the user cannot: • Post new content • Send messages • Follow new accounts • Create Groups • Interact with content (like, reply, repost)

Suspension durations: • 12 hours (minor violations, first offense) • 24 hours (moderate violations) • 7 days (serious violations) • 30 days (severe violations, pending review)

B. Termination The cryptographically enforced death of the Node: • The Public Key is blacklisted (cannot create new Node with same identity) • The Device Fingerprint is banned (cannot create new Node on same device) • The IP Hash is flagged (new accounts from same IP monitored closely) • All Public Content is removed • Encrypted Private Content becomes inaccessible (keys destroyed)

Termination is permanent and irreversible. Terminated Nodes cannot be appealed except in cases of proven error (e.g., false positive on CSAM detection).

C. Immediate Termination Offenses The following result in instant Termination without warning: • CSAM (Section 2.1) • Terrorism recruitment or material support (Section 1.1) • Swatting (Section 2.4) • Severe doxxing with credible threat to life (Section 2.4) • Sale of narcotics or weapons (Section 1.3)

Section 6.4: Evidence Preservation

For violations involving illegal content (CSAM, terrorism, severe harassment), we preserve: • IP Hashes (pausing rotation cycle) • Timestamps of all activity • Device fingerprints • Public posts and metadata • Report details

This data is moved to forensic cold-storage and may be provided to law enforcement upon valid legal process or emergency disclosure.

ARTICLE VII: APPEALS & OVERSIGHT

Kemet acknowledges that algorithms are imperfect and that power requires accountability. We provide paths to justice.

Section 7.1: The Appeal Process

A. Submission Users may appeal a Suspension or Termination via: • In-app appeal form (Settings > Account > Appeal) • Email to appeals@kemet.network

B. Requirements The appeal must include: • Node ID • Specific enforcement action being appealed (Shadow, Suspension, Termination) • Written statement defending the content or explaining the context • Any evidence supporting the appeal (screenshots, context, etc.)

C. Review Appeals are reviewed by the Oversight Team—human administrators trained in the Common Sense Doctrine. Automated systems flag cases; humans make decisions.

D. Timeline • Acknowledgment: Within 48 hours • Initial review: Within 7 days • Final decision: Within 30 days (complex cases may take longer)

E. Outcome The Oversight Team may: • Uphold the enforcement action (appeal denied) • Modify the enforcement (reduce Suspension duration, convert Termination to Suspension) • Reverse the enforcement (appeal granted, content restored, Node reinstated) • Request additional information

Section 7.2: The Standard of Review

A. Context Matters We do not ban words; we ban intent. If a user can prove their content was: • Quoted for educational purposes • Satirical or parodic (and clearly labeled) • Misidentified by automated systems (false positive) • Shared without understanding context (naive repost)

Then the ban may be overturned.

B. Finality Decisions made by the Oversight Team regarding the following are FINAL and not subject to further appeal: • CSAM (Section 2.1) — mandatory reporting to law enforcement • Terrorism (Section 1.1) — mandatory law enforcement handoff • Severe doxxing with credible threat (Section 2.4)

These decisions involve legal obligations that supersede individual appeal rights.

C. Transparency All administrative actions are logged in an immutable, hash-chained audit trail (Blockchain). This includes: • Timestamp of action • Type of action (Shadow, Suspension, Termination) • Reason code • Administering party identifier (hashed)

We may publish aggregate statistics about enforcement actions in periodic transparency reports.

Section 7.3: The "Scorched Earth" Clause

If a Node is terminated for severe illegal content (CSAM, Terrorism, life-threatening criminal activity), Kemet reserves the right to:

1. Retain encrypted data for law enforcement (as per Privacy Policy) 2. Report the user to ISP and local authorities (where legally permitted) 3. Federate the ban across partner safety networks (e.g., GIFCT hash sharing for CSAM) 4. Preserve evidence for potential criminal prosecution

This clause applies only to the most severe violations that threaten life, safety, or The Network itself.

ARTICLE VIII: AMENDMENTS & EVOLUTION

Section 8.1: Living Document

This Covenant is a living document. As new threats emerge—AI-generated biological weapon blueprints, novel forms of deepfake exploitation, quantum decryption risks—Kemet reserves the right to amend these standards.

Section 8.2: Notification of Changes

Significant changes to the "Red Lines" (Zero Tolerance policies) will be broadcast via: • System Notification to all active Nodes • Email to registered support addresses (if provided) • Posting on legal.kemet.network • In-app announcement

Users will have 30 days to review changes before they take effect.

Section 8.3: Retroactivity

New rules apply forward, not backward. However, egregious historical violations that pose an ongoing threat to The Network may be actioned if they were violations of the original Covenant and remain violations of the amended Covenant.

Section 8.4: Community Input

We may solicit community input on proposed changes through: • Public comment periods • Surveys of active users • Advisory councils

However, final decisions rest with Kemet Protocol Governance. Safety is not a democracy.

APPENDIX A: DEFINITIONS AND INTERPRETATION

For purposes of this Covenant:

"Content" means any information, data, text, images, video, audio, or other material transmitted through The Network.

"CSAM" means Child Sexual Abuse Material, as defined in Section 2.1.

"Device Fingerprint" means a hashed representation of device characteristics used for security.

"Direct Message" or "DM" means end-to-end encrypted private communication between Nodes.

"Doxxing" means publishing private information with intent to facilitate harassment.

"E2EE" means end-to-end encryption.

"Group" means a collection of Nodes for shared communication, which may be public (discoverable) or private (encrypted, invite-only).

"HED Exception" means Historical, Educational, or Documentary context.

"IP Hash" means a salted, hashed representation of an IP address.

"Node" means a user account and its associated cryptographic identity.

"Node Termination" means permanent disabling of a Node and blacklisting of associated identifiers.

"NSFW" means Not Safe For Work—content that requires viewer discretion.

"Public Content" means content posted with visibility set to "public" or shared in discoverable Groups.

"Public Feed" means the algorithmically and chronologically ordered display of public content.

"Visibility Restriction" means the enforcement state where a Node's public content is de-indexed and removed from global feeds, with notification provided to human operators.

"Swatting" means falsely reporting emergencies to send armed responders to a victim's location.

"Trust Score" means the algorithmic reputation metric assigned to each Node.

"Wikilawyering" means attempting to bypass rules through technicalities or legalistic arguments.

APPENDIX B: ENFORCEMENT REFERENCE TABLE