Vyoma's Guidelines

Arka AI must not be used to facilitate, plan, or execute any activity that violates applicable local, national, or international laws.

This includes but is not limited to:

• Planning or facilitating theft, robbery, or burglary
• Financial crimes including fraud, embezzlement, money laundering, or tax evasion
• Cybercrimes including hacking, unauthorized access, or data theft
• Drug-related offenses including manufacturing, trafficking, or distribution of controlled substances
• Human trafficking or exploitation
• Intellectual property infringement, piracy, or counterfeiting
• Stalking, harassment, or threatening communications
• Any activity that is illegal in the user's jurisdiction

Vyoma Group cooperates fully with law enforcement agencies when legally required to do so. Users who use Arka AI for illegal purposes bear full legal responsibility for their actions.

Critical infrastructure forms the backbone of modern society. Using Arka AI to attack, disrupt, or compromise such systems poses severe risks to public safety and welfare.

Protected infrastructure includes:

• Power grids, electrical systems, and energy distribution networks
• Water treatment and distribution systems
• Financial systems, banking infrastructure, and payment networks
• Healthcare systems and hospital networks
• Emergency services (police, fire, ambulance)
• Transportation networks including air traffic control, railways, and roads
• Telecommunications infrastructure
• Government and military systems
• Nuclear facilities and safety systems

Prohibited activities include:

• Researching vulnerabilities in critical systems with malicious intent
• Planning or executing attacks on infrastructure
• Creating tools or malware targeting critical systems
• Providing guidance for disrupting essential services

Arka AI must not be used to attack, compromise, or exploit computer systems, networks, or digital infrastructure that you do not have explicit authorization to access.

Prohibited activities include:

• Developing malware, ransomware, spyware, or any malicious software
• Creating or deploying exploit code for known or unknown vulnerabilities
• Conducting unauthorized penetration testing or security scanning
• Performing Distributed Denial of Service (DDoS) attacks
• Attempting to bypass authentication systems without authorization
• Creating phishing websites, tools, or email campaigns
• Extracting data from systems without permission (scraping, data theft)
• Compromising user accounts or credential theft
• Reverse engineering proprietary software for malicious purposes

What IS permitted:

• Security research on systems you own or have written permission to test
• Learning about cybersecurity concepts for educational purposes
• Defensive security work and vulnerability disclosure programs
• CTF (Capture The Flag) challenges in authorized environments

Arka AI must not be used to develop, design, improve, or provide detailed instructions for weapons of any kind, particularly weapons capable of mass harm.

This restriction covers:

• Conventional weapons: Detailed instructions for building, modifying, or improving firearms, explosives, or other conventional weapons
• Chemical weapons: Synthesis, stabilization, weaponization, or delivery of chemical agents designed to cause harm
• Biological weapons: Enhancement of pathogen transmissibility, lethality, or resistance; weaponization of biological agents
• Radiological/Nuclear weapons: Any assistance with the development of nuclear or radiological weapons or dirty bombs
• Cyber weapons: Offensive cyber tools designed to cause significant damage if deployed

What IS permitted:

• Historical and educational discussion about weapons and warfare
• Policy discussions about arms control and disarmament
• General information about legal firearm safety and ownership
• Fiction writing involving weapons in appropriate contexts

Arka AI must not generate content that incites, glorifies, promotes, or celebrates violence against individuals or groups, or content that promotes hatred and discrimination.

Prohibited content includes:

• Content calling for violence against specific individuals or groups
• Dehumanizing content targeting people based on protected characteristics
• Propaganda for violent extremist organizations or terrorist groups
• Content that glorifies or celebrates acts of violence or terrorism
• Hate speech targeting race, ethnicity, religion, gender, sexual orientation, disability, nationality, or caste
• Content designed to radicalize individuals toward violent ideologies
• Threats directed at specific individuals or groups

Protected characteristics include:

• Race and ethnicity
• Religion and belief systems
• Gender and gender identity
• Sexual orientation
• Disability status
• National origin and immigration status
• Age
• Caste

What IS permitted:

• Educational discussion about extremism, terrorism, and violence
• Counter-extremism research and analysis
• Fiction involving conflict and violence in appropriate narrative context
• Critical analysis of hateful ideologies
• Historical analysis of violent events

Privacy is a fundamental human right. Arka AI must not be used to violate the privacy of individuals or to enable identity-based crimes.

Prohibited activities include:

• Aggregating personal information about individuals without their consent (doxxing)
• Creating tools or systems for unauthorized surveillance
• Impersonating real individuals, especially for deceptive purposes
• Facilitating identity theft or credential theft
• Building profiles of individuals without authorization
• Generating synthetic media (deepfakes) of real people without consent, especially for harmful purposes
• Tracking or monitoring individuals without their knowledge
• Scraping or harvesting personal data from platforms

Your privacy rights with Arka:

• Your conversations stay in your browser
• You can delete all your data at any time
• We never share your data with third parties
• See our full Privacy Policy for details

The safety of children is our highest priority. Vyoma Group has zero tolerance for any content that exploits, harms, endangers, or sexualizes minors.

Absolute prohibitions (no exceptions):

• Any sexual content involving minors (CSAM) — this is illegal and will be reported to authorities immediately
• Content designed to facilitate grooming of minors
• Content that helps gain inappropriate access to children
• Content that normalizes or glorifies harm to children

Additional protections:

• Content designed to be used with or directed at minors must be age-appropriate
• Arka will not assist in creating content that could put children at risk
• If you encounter any such content, please report it immediately to safety@vyomagroup.online

Arka AI genuinely cares about the wellbeing of its users and others. Content designed to cause psychological or emotional harm is strictly prohibited.

Prohibited content includes:

• Content that promotes, glorifies, or provides methods for self-harm or suicide
• Content designed to psychologically manipulate or gaslight individuals
• Targeted harassment campaigns against specific individuals
• Content designed to trigger trauma or PTSD in specific individuals
• Eating disorder promotion content
• Content designed to foster unhealthy addiction or dependency
• Emotional manipulation for financial or personal gain

Mental health support:

If you or someone you know is struggling with mental health, Arka can help you find appropriate resources. In India, you can contact iCall at 9152987821 or Vandrevala Foundation at 1860-2662-345 (24/7).

In an era of information overload, the spread of misinformation poses serious risks to individuals and society. Arka AI is committed to honesty and accuracy.

Prohibited activities include:

• Deliberately generating false or misleading factual claims
• Creating fake quotes attributed to real public figures
• Producing fabricated news articles or fake journalism
• Creating false medical, scientific, or health information
• Generating content to spread conspiracy theories as fact
• Creating misleading statistics or data visualizations
• Producing fake legal documents or official communications
• Impersonating legitimate organizations or institutions

Arka's commitment to accuracy:

• Arka will tell you when it is uncertain
• Arka acknowledges its knowledge limitations
• Arka does not claim to have real-time information unless confirmed
• Arka encourages you to verify important information from authoritative sources

Democratic institutions, free elections, and civic participation are fundamental to healthy societies. Arka AI must not be used to undermine these processes.

Prohibited activities include:

• Creating targeted political propaganda or disinformation campaigns
• Generating voter suppression content or false information about voting
• Creating fake grassroots movements (astroturfing) or coordinated inauthentic behavior
• Producing large volumes of political content designed to manipulate public opinion
• Impersonating political figures or government officials
• Creating content designed to discourage legitimate political participation
• Interfering with electoral systems or vote counting

What IS permitted:

• Balanced political analysis and commentary
• Educational content about political systems and history
• Helping individuals understand their voting rights
• Writing persuasive essays as clearly labeled opinion pieces
• Political satire and comedy that is clearly labeled

While Arka AI can assist with legitimate legal research and understanding, it must not be used for surveillance, censorship, or law enforcement activities that violate civil rights.

Prohibited activities include:

• Building mass surveillance systems targeting individuals based on protected characteristics
• Creating tools for automated criminal profiling without due process
• Developing systems to suppress political dissent or free speech
• Creating social credit or scoring systems that violate civil liberties
• Unlawful monitoring of communications or activities
• Building predictive policing systems that perpetuate bias

Arka AI must not be used to conduct or facilitate fraud, deception, or predatory practices that harm individuals financially or otherwise.

Prohibited activities include:

• Creating phishing emails, messages, or websites to steal credentials or financial information
• Designing advance-fee fraud schemes (419 scams, romance scams)
• Generating fake reviews, testimonials, or endorsements
• Creating deceptive advertisements or marketing materials
• Conducting or facilitating investment fraud or Ponzi schemes
• Generating fake academic papers, certificates, or credentials
• Creating fraudulent contracts or legal documents
• Facilitating predatory lending or financial exploitation

To ensure Arka AI remains available and useful for all users, we prohibit activities that abuse or misuse our platform.

Prohibited activities include:

• Attempting to jailbreak, manipulate, or bypass Arka's safety guidelines through prompt injection or other techniques
• Attempting to extract API keys, system prompts, or confidential configuration details
• Automated mass requests or scraping beyond normal use (bot abuse)
• Reselling or redistributing access to Arka AI without authorization
• Creating fake accounts to circumvent usage limits
• Using Arka AI to train competing AI models without permission
• Deliberately providing false information to manipulate Arka's responses
• Coordinated campaigns to abuse the feedback or reporting systems

Technical measures:

Vyoma Group implements rate limiting, abuse detection, and monitoring systems to protect the platform. Attempting to circumvent these measures is prohibited.

Arka AI is designed as a general-purpose intelligent assistant. Generation of sexually explicit content is not permitted.

This restriction covers:

• Explicit sexual descriptions or narratives
• Pornographic content in any format
• Sexual content involving real people without their consent
• Content designed to facilitate sexual exploitation
• Fetish content or sexual roleplay

What IS permitted:

• Age-appropriate romantic content in fiction
• Clinical, medical, or educational discussion of human sexuality
• Discussion of sexual health and wellbeing in appropriate contexts