Security & Privacy Architecture. Rigorous safeguards for confidential AI workflows.
An exhaustive breakdown of ChatBox AI's defense-in-depth model: contractual zero model training guarantees, in-transit and at-rest cryptographic isolation, hashed developer key management, and sovereign compliance under the Indian DPDP Act.
Zero Model Training TLS 1.3 & AES-256 PCI-DSS Level 1 Isolation Primary Hosting: India Developer: Arpit Ariyan Mahanarana (technon.co.in)DPDP Act & GDPR Aligned
Core Security Pillars
6 Architectural Safeguards
01Contractual Policy
Zero Public Model Training
Your prompts, code snippets, project briefs, and generated websites are never used to train, refine, or fine-tune public foundation models. We enforce strict enterprise zero-retention API contracts with OpenAI, Anthropic, and Google AI.
No prompt data leakage into public foundation weights
No evaluation or reinforcement training on customer inputs
Full intellectual property ownership over generated outputs
02End-to-End
Cryptographic Data Protection
All communications between your browser and our servers run strictly over TLS 1.3 with modern cipher suites. Database volumes, account identifiers, and conversation histories are encrypted at rest using industry-standard AES-256.
TLS 1.3 transport encryption with Strict Transport Security (HSTS)
AES-256 encrypted storage volumes across databases
Ephemeral memory processing for real-time model inferences
03Access Hardening
Multi-Factor Authentication & Identity
We support email-based One-Time Password (OTP) Multi-Factor Authentication to shield accounts against automated password stuffing and session hijack attempts.
MFA available directly in Account Settings
Strict JSON Web Token (JWT) session invalidation on logout
Automated rate-limiting and IP throttling against brute-force attacks
04Developer Safety
Hashed API Key Architecture
Developer API keys created for our OpenAI-compatible endpoint are stored strictly as salted SHA-256 hashes. Even in the event of an internal database audit, plaintext keys cannot be decrypted or stolen.
One-time display of secret keys upon initial creation
Salted SHA-256 cryptographic hashing in database tables
Instant one-click key revocation and rotation in Settings
05PCI-DSS Level 1
Payment & Financial Isolation
ChatBox AI servers never process, collect, or store raw credit card numbers, CVV codes, or net-banking passwords. All payment transactions are securely offloaded to certified payment partners.
Domestic Indian transactions (UPI, RuPay, Netbanking) via Razorpay
International subscription management via Paddle
Only tokenized transaction identifiers and invoice IDs are saved
06DPDP Act & GDPR
Data Sovereignty & Right to Erasure
In compliance with India’s Digital Personal Data Protection (DPDP) Act, users retain absolute sovereignty over their data. You can delete individual conversation threads or execute a full account purge at any time.
Primary database clusters hosted in Indian datacenter regions
Permanent wipe of profile records and threads within 30 days of deletion
Invoice records retained for 7 years per Indian corporate taxation laws
Security & Isolation Matrix
Across Tiers
Protection Feature
Free Plan
Pro Plan (₹499/mo)
Max Plan (₹1,999/mo)
Zero Public Model Training
Included
Included
Included
TLS 1.3 & AES-256 Volume Encryption
Included
Included
Included
Multi-Factor Authentication (MFA)
Included
Included
Included
Developer API Keys (Hashed SHA-256)
No
Included (60 req/min)
Included (120 req/min)
Data Deletion & 30-Day Purge SLA
Included
Included
Included
Support Response Priority
Standard (48–72h)
Priority SLA (< 24h)
Dedicated Engineering SLA
Sub-Processors & Data Sovereignty
Verified Partners
ChatBox AI strictly limits data transmission to vetted enterprise cloud providers. We never share, sell, or syndicate user data with data brokers, ad networks, or telemetry trackers.
Provider Entity
Role & Data Processed
Datacenter Location
Cryptographic Standard
Firebase (Google)
Authentication & Session Token Verification
India / Global
Encrypted JWTs, OAuth 2.0
Supabase & Appwrite
Database & Private File Storage
India / Frankfurt
AES-256 Volume Encryption, Row Level Security
Razorpay & Paddle
Payment Processing & Subscription Management
India / UK
PCI-DSS Level 1 Certified, Tokenized Vaults
OpenAI, Anthropic, Google AI
Foundation Model Inference Providers
US / EU / Global
Commercial Zero-Training Enterprise APIs
Recommended User Security Practices
Shared Responsibility
Enable MFA
Turn on email OTP verification in Settings to prevent credential-stuffing attacks.
Protect API Keys
Never commit developer API keys into public Git repositories or client-side web bundles.
Sanitize Secrets
Avoid pasting production passwords, live tokens, or confidential PII into prompt text.
Purge Stale Threads
Regularly delete completed project conversations and website drafts from your dashboard.
Responsible Vulnerability Disclosure
Report Security Vulnerabilities
We welcome reports from independent security researchers. If you identify a potential security vulnerability, report it to our Data Protection Officer and security engineering team for prompt investigation and triage.