Acceptable Use and AI Policy

1.1 PURPOSE

This Acceptable Use Policy ("Policy") sets forth the standards of conduct that users must follow when using Algorithec's Platform. This Policy applies to all users, including free users, paying customers, and partners.

1.2 PROHIBITED USES

A. Illegal Activities

Users must NOT:

• Violate any federal, state, local, or international laws
• Engage in fraud, deception, or forgery
• Facilitate money laundering or terrorist financing
• Evade taxes or misrepresent income
• Engage in cybercrime or hacking
• Violate consumer protection laws
• Facilitate unlicensed financial services
• Engage in any form of trafficking
• Facilitate unauthorized distribution of controlled substances

Consequences:

• Immediate account termination
• Data preservation for law enforcement
• Legal action and criminal referral
• Cooperation with authorities

B. Harassment & Abuse

Users must NOT:

• Harass, stalk, or threaten other users
• Engage in cyberbullying or targeted harassment
• Create hostile environment for protected classes
• Discriminate based on race, color, religion, caste, sex, gender identity, sexual orientation, national origin, age, or disability
• Defame, slander, or libel others
• Reveal private information without consent (doxxing)
• Engage in sexual harassment
• Make death threats or threats of violence

Consequences:

• Warning for first offense (usually)
• Account suspension for escalation
• Permanent account termination
• Potential legal action
• Law enforcement notification (if warranted)

C. Platform Manipulation & Abuse

Users must NOT:

• Attempt to hack or gain unauthorized access
• Reverse engineer or decompile code
• Use bots, scrapers, or automated tools
• Perform DDoS or DoS attacks
• Create multiple accounts to bypass restrictions
• Manipulate AI algorithm through fake reviews
• Exploit security vulnerabilities
• Access others' accounts without permission
• Use proxy servers to bypass restrictions
• Interfere with Platform operations

Consequences:

• Immediate account termination
• IP ban
• Legal action for damages
• Cooperation with law enforcement

D. Spam & Unwanted Communications

Users must NOT:

• Send unsolicited commercial messages
• Send phishing or malware
• Engage in network marketing or pyramid schemes
• Send chain letters or mass emails
• Advertise unauthorized products/services
• Send misleading or deceptive messages
• Flood Platform with repetitive content
• Engage in email spoofing

Consequences:

• Content removal
• Account suspension
• Permanent termination
• IP ban

E. Intellectual Property Violations

Users must NOT:

• Infringe copyrights (sharing pirated content, unauthorized copying)
• Violate trademarks or service marks
• Plagiarize others' content
• Claim ownership of others' work
• Violate patents
• Violate trade secrets
• Distribute copyright-protected material
• Create derivative works without permission

Consequences:

• Content removal
• Account suspension
• DMCA takedown compliance
• Legal action for damages
• Permanent account termination

F. Fraudulent Activity

Users must NOT:

• Make false claims about products/services
• Misrepresent themselves or affiliations
• Provide false information during registration
• Engage in chargebacks for legitimate purchases
• Return stolen goods as legitimate returns
• Create fake reviews or manipulate ratings
• Engage in payment fraud
• Use stolen payment methods
• Impersonate Algorithec staff
• Engage in advance-fee fraud

Consequences:

• Immediate account termination
• Legal action
• Criminal referral to authorities
• Cooperation with law enforcement
• Blacklist to prevent re-registration

G. Adult & Explicit Content

Users must NOT:

• Upload or share sexually explicit content
• Share child sexual abuse material (illegal, serious consequences)
• Share content involving minors in adult situations
• Share non-consensual intimate images ("revenge porn")
• Engage in sex trafficking or exploitation
• Solicit sexual services
• Create adult content involving real people without consent

Consequences:

• Immediate account termination
• Content removal and preservation
• Criminal referral to law enforcement
• Blacklist and ban
• Law enforcement cooperation

H. Violence & Dangerous Content

Users must NOT:

• Share content depicting graphic violence or gore
• Provide instructions for self-harm
• Provide instructions for creating weapons
• Glorify or encourage violence
• Provide instructions for illegal drugs
• Share content promoting terrorism
• Provide suicide instructions
• Encourage dangerous activities

Consequences:

• Content removal
• Account suspension
• Permanent termination
• Law enforcement notification (if warranted)
• Crisis intervention resources provided

1.3 ALGORITHMIC INTEGRITY & AI FAIRNESS

A. Fair Use of AI System

Users must:

• Accept AI recommendations in good faith
• Not attempt to manipulate AI decisions
• Provide accurate preferences and instructions
• Not create fake reviews to game AI
• Accept that AI decisions are automated
• Verify recommendations before accepting
• Report AI errors or bias

Users must NOT:

• Attempt to reverse engineer AI
• Try to consistently reject AI recommendations to game system
• Create multiple accounts to manipulate AI training
• Provide false information to train AI models
• Attempt to create bias in AI through coordinated fake reviews
• Use AI for unauthorized bulk transactions
• Resell AI decision service to others

B. Transparency & Disclaimer Acceptance

Users acknowledge:

• AI decisions are automated, not human-reviewed
• AI bases decisions on available data (may be outdated)
• AI may not capture all preferences
• Prices and availability change after AI decision
• ONDC/partner policies still apply
• Final decision remains user's responsibility
• AI errors and limitations are possible
• User should verify before accepting

1.4 COMMERCE PARTNER VIOLATIONS

A. ONDC Seller Fraud

Users must NOT:

• Use Algorithec to defraud ONDC sellers
• Use fake payment methods for ONDC purchases
• Request refunds for items received and used
• Claim items not received when they were
• Return counterfeit items as originals
• Engage in chargeback fraud
• Coordinate refund fraud with other users

B. Rating & Review Manipulation

Users must NOT:

• Pay for positive reviews
• Accept payment for positive reviews
• Vote on reviews based on competitor relationships
• Create fake accounts to review
• Leave reviews for products not purchased
• Leave reviews for false experiences
• Coordinate review manipulation
• Leave reviews unrelated to product

ONDC & Partner Consequences:

• Review removal
• Account restrictions on ONDC
• Seller/buyer blacklist
• Potential legal action by seller

1.5 PRIVACY & DATA MISUSE

Users must NOT:

• Collect other users' personal data
• Scrape the Platform
• Use data for unauthorized purposes
• Share others' personal information
• Engage in unauthorized data mining
• Violate others' privacy expectations
• Access others' accounts

1.6 ENFORCEMENT & CONSEQUENCES

Tiered Approach:

Warning (First Offense - Usually):

• Violation notice
• Description of violation
• Opportunity to cure
• Policy clarification
• Account remains active

Suspension (Second Offense or Severity):

• Temporary access loss (7-30 days)
• Inability to make purchases
• Ability to access data
• Right to appeal
• Clear timeline for reinstatement
• May require corrective action

Permanent Termination (Repeated or Severe):

• Immediate account loss
• No refunds for unused credits
• Blacklist to prevent re-registration
• Potential legal action
• Data preservation for authorities
• No reinstatement option

Immediate Termination (Serious Violations):

• Illegal activity
• Child exploitation
• Terrorism or violence
• Extreme harassment
• Fraud or repeated fraud
• System attacks

PART 2: AI DECISION ENGINE SPECIFIC POLICY

2.1 HOW THE AI WORKS

Our AI Decision Engine:

• Understands Intent - NLP processes your natural language request
• Gathers Data - Searches across ONDC, partner platforms in real-time
• Evaluates Options - Scores based on multiple factors:
• Price (lowest cost option)
• Delivery speed (fastest option)
• Reliability (seller/service rating)
• Quality signals (user reviews, return rates)
• Discounts/offers (current promotions)
• Your historical preferences (learning model)
• Budget constraints (your specified limits)
• Time constraints (your urgency level)

• Ranks Results - Proprietary algorithm weights factors
• Applies Discounts - Auto-applies available coupons, cashback
• Makes Decision - Presents best option for acceptance
• Executes Transaction - Places order once you confirm
• Tracks Fulfillment - Updates you on status

2.2 AI LIMITATIONS YOU SHOULD KNOW

The AI Cannot:

• Guarantee lowest possible price (real-time changes)
• Predict future price drops (incomplete market data)
• Guarantee seller reliability (rating-based estimation)
• Know all available options (limited to integrated platforms)
• Capture all your preferences (incomplete user input)
• Predict quality (based on aggregate reviews)
• Guarantee availability (changes constantly)
• Account for all delivery options (depends on partner availability)
• Know future promotions (based on current data)
• Read between the lines (based on stated intent)

External Factors AI Can't Control:

• Seller inventory changes
• Price changes between decision and purchase
• Discount expiration
• Platform outages
• Delivery delays
• Quality issues with seller
• Shipping damages
• Returns/refund policies
• Payment issues
• Network problems

2.3 AI TRANSPARENCY & EXPLAINABILITY

What We Show You:

• Top 3 options considered
• Why we ranked #1 option first
• Key factors in decision
• Price comparison
• Delivery estimates
• Seller ratings
• Available discounts

What We Don't Show:

• Exact algorithm weights
• Training data composition
• Complete decision tree
• Proprietary ranking formula
• Full option evaluation (too many to display)

2.4 AI TRAINING & IMPROVEMENTS

How We Improve AI:

• Analyze decision patterns
• Study user feedback
• Learn from corrections
• Identify errors/gaps
• Update models regularly
• Test new algorithms
• A/B test improvements
• Measure accuracy metrics

Your Data & AI Training:

• We use anonymized data to improve AI
• You can opt-out of AI training (Settings Privacy)
• We don't identify individuals in training
• Individual decisions not stored as examples
• Aggregate patterns used for learning
• Zero-party data (stated preferences) used
• Behavioral signals analyzed

AI Fairness:

• We monitor for bias
• Test for discrimination
• Regular audits conducted
• Address disparities found
• Transparent about limitations
• Accept feedback on bias
• Continuously improve fairness

2.5 AI DECISION APPEALS

If You Disagree With AI Decision:

• Select "Why this option?"
• Provide feedback on factors
• Specify what you wanted instead
• Submit improvement suggestion
• Our team reviews feedback
• We retrain models based on feedback

Example Appeals:

• "I wanted fastest delivery, not cheapest"
• "This seller has bad reviews, should rank lower"
• "You missed option X from my preferred seller"
• "Price comparison seems wrong"
• "This option is no longer available"

Resolution:

• Review within 5 business days
• Corrective action if needed
• Model improvement if applicable
• Response provided to user
• Appeal tracking for AI improvement

2.6 AUTOMATED DECISIONS & YOUR RIGHTS

Your Rights Regarding Automated Decisions:

Under Indian law, you have the right to:

• Know a decision was automated
• Request human review of automated decision
• Understand factors used in decision
• Contest the decision
• Appeal to human decision-maker

Our Commitment:

• Be transparent that AI made decision
• Explain factors considered
• Allow human override
• Accept your manual overrides
• Not force automated decisions
• Provide appeals process
• Maintain human support option

You Can Always:

• Ask for human assistant
• Manually select option from list
• Reject AI recommendation
• Contact support for manual help
• Request human decision-maker review

2.7 AI ERROR HANDLING

If AI Makes Error:

• Report immediately via support
• Provide error description
• Include order/transaction ID
• We investigate within 24 hours
• Escalate to AI team
• Provide resolution/correction
• Adjust recommendation if needed
• Improve model to prevent recurrence

Common Errors:

• Price comparison inaccuracy
• Seller rating mismatch
• Availability not updated
• Discount already expired
• Wrong category selection
• Delivery estimate wrong
• Item not as described by AI

Our Response:

• Acknowledge error
• Correct if possible
• Offer alternative
• Provide compensation if appropriate
• Prevent similar errors
• Improve models

2.8 AI BIAS & FAIRNESS

We Commit To:

• Regular bias audits
• Testing for discrimination
• Monitoring for disparate impact
• Addressing identified biases
• Transparent about limitations
• Accepting feedback on bias
• Continuous fairness improvement

Prohibited AI Use:

• Discrimination in recommendations
• Excluding certain groups
• Giving inferior options to protected classes
• Pricing discrimination
• Fraud detection with discriminatory bias

If You Experience Bias:

• Report to: ai-fairness@algorithec.ai
• Describe the issue
• Provide examples
• We investigate
• Corrective action taken
• You notified of resolution

PART 3: ONDC INTEGRATION POLICY

3.1 ONDC NEUTRALITY

Algorithec Remains Neutral:

• No preference for any ONDC seller
• Rankings based on objective factors only
• No paid placement or preferential treatment
• Same evaluation criteria for all sellers
• No conflicts of interest
• Transparent ranking factors
• Appeal process available

What We Don't Do:

• Accept payment for better rankings
• Give preferential treatment to partners
• Hide negative reviews
• Manipulate ratings
• Exclude sellers arbitrarily
• Show bias toward certain sellers

3.2 SELLER DATA & INFORMATION

Information We Collect About Sellers:

• Product catalog (name, description, price)
• Ratings and reviews
• Response times
• Fulfillment metrics
• Return/refund rates
• Policy compliance
• Availability information
• Promotional offers

Seller Data Use:

• Ranking and evaluation
• Quality metrics
• Reliability assessment
• Recommendation training
• Performance analytics
• No sharing with competitors
• Confidential treatment
• ONDC data agreement compliance

3.3 SELLER DISPUTES & COMPLAINTS

For Seller Issues:

• Contact seller first (via ONDC)
• Algorithec facilitates, doesn't adjudicate
• Seller policies apply
• ONDC arbitration available
• Algorithec not liable for seller disputes
• We provide contact information
• We facilitate communication

If Seller Violates Law:

• Report to ONDC
• We preserve evidence
• Cooperate with ONDC investigation
• May limit seller access (if warranted)
• Support to affected users

3.4 PLATFORM INTEGRATION REQUIREMENTS

Algorithec Guarantees:

• Real-time price sync (within 5 minutes)
• Inventory accuracy
• Current discounts
• Delivery estimates
• Order placement confirmation
• Status tracking integration
• Return/refund processing
• Customer support liaison

PART 4: DATA SECURITY IN AI CONTEXT

4.1 AI-SPECIFIC DATA PROTECTION

Data Used for AI:

• Anonymized decision patterns
• Aggregate user preferences
• Aggregated behavior signals
• Performance metrics
• No individual identification

Data NOT Used for AI:

• Personal identifiable information
• Payment details
• Contact information
• Location history (only current location)
• Medical/sensitive information
• Behavioral targeting without consent

Your Control:

• Opt-out of AI training
• Settings Privacy AI Training
• Effective immediately
• May reduce recommendation quality
• No penalty for opting out

4.2 MODEL SECURITY

We Protect:

• AI models from theft
• Training data from unauthorized access
• Algorithm from reverse engineering
• Proprietary processes
• Trade secrets

We Don't:

• Share models with third parties
• Make models publicly available
• Disclose training procedures
• Reveal ranking formulas
• Expose proprietary data

PART 5: COMPLIANCE & ENFORCEMENT

5.1 INVESTIGATION PROCESS

For Violations:

• Detection - Automated systems and user reports
• Notification - User informed of alleged violation
• Investigation - Evidence gathered, reviewed
• Review - Violation confirmed or dismissed
• Response - Warning, suspension, or termination
• Appeal - User can appeal within 30 days
• Resolution - Final decision provided

5.2 APPEAL PROCESS

To Appeal:

• Email: appeals@algorithec.ai
• Provide order/transaction ID
• Explain disagreement
• Provide supporting evidence
• Submit within 30 days

Appeal Review:

• Independent reviewer (not original decision maker)
• Full context considered
• User statement reviewed
• Evidence evaluated
• Decision upheld, reversed, or modified
• Final response within 30 days

5.3 RECORD RETENTION

We retain records of:

• Violations and enforcement actions
• Appeals and outcomes
• Evidence (90 days minimum)
• Aggregate violation data (for trends)
• User warnings and suspensions

Your Access:

• Request your enforcement record
• Email: support@algorithec.ai
• Provided within 10 business days
• Subject to privacy protections