AI & Data Processing Policy
Effective 27 August 2024. Last updated 25 September 2026.
Preamble
This AI & Data Processing Governance Policy ("Policy") constitutes a binding, framework governing the design, development, deployment, operation, limitation, and legal positioning of artificial intelligence systems and data processing activities conducted by INSPYR FINNOVATION PRIVATE LIMITED, operating under the brand name "Senseible" ("Senseible", "Company").
This Policy is drafted to withstand scrutiny from regulators, financial institutions, enterprise customers, auditors, investors, and courts across India, the European Union, and emerging markets. It supersedes and replaces all prior AI‑related disclosures, summaries, or statements.
0. Legal Characterisation of the Platform
Senseible operates as a digital climate‑intelligence and data‑processing infrastructure provider. Senseible does not operate as an autonomous decision‑making system. All AI systems deployed by Senseible function as assistive analytical tools that generate probabilistic outputs based on statistical inference.
Senseible expressly disclaims classification as a regulated decision‑making authority, advisor, verifier, auditor, or certifying body under any jurisdiction.
1. Purpose & Scope
Senseible operates an AI-native climate infrastructure platform that transforms business data into carbon intelligence, ESG signals, and climate-finance enablement. This Policy exists to ensure transparency, accountability, and legal clarity around AI usage, data processing, and model governance.
This Policy establishes enforceable rules to: (a) allocate risk between Senseible and counterparties; (b) define lawful purposes and boundaries of AI use; (c) preserve the integrity of proprietary models; (d) prevent regulatory misclassification; and (e) mitigate systemic, operational, and model‑related risks.
2. Classification of AI Systems
Senseible deploys multiple AI systems including:
- Optical Character Recognition (OCR) for invoice and document ingestion
- Large Language Models (LLMs) for data normalization, categorization, and inference
- Emission Foundation Models for Scope 1–3 estimation
- Statistical anomaly, fraud, and risk‑detection systems
- Voice-based AI systems for MSME onboarding and interaction
All AI systems are assistive and inferential in nature and do not constitute decision-making authorities.
3. Nature of AI Outputs
AI-generated outputs are probabilistic estimates based on statistical modeling, historical datasets, and inferred patterns.
Outputs are not deterministic facts, certified measurements, or regulatory filings.
Variability may arise due to data quality, model evolution, regulatory changes, or third-party dependencies.
4. Human Oversight & Responsibility
Senseible maintains human oversight over system design, monitoring, and improvement.
Users remain solely responsible for interpreting outputs and ensuring compliance with applicable laws, audits, and disclosures.
5. Data Inputs & Quality Dependency
AI outputs are materially dependent on the quality, completeness, timeliness, and accuracy of input data. Senseible does not independently authenticate, audit, or verify raw data unless expressly agreed under a separate written instrument.
6. Model Training & Learning
Senseible uses user data to train, fine-tune, and improve proprietary AI models.
Training uses aggregated, anonymized, and non-identifiable datasets wherever feasible.
Users may opt out of future training use via written request.
Learnings already embedded into models cannot be reversed or extracted.
7. Bias, Error & Drift
AI systems may reflect biases present in training data.
Models may drift over time due to changing regulations, markets, or data patterns.
Senseible continuously monitors performance but does not guarantee bias-free or error-free outputs.
8. Prohibited Uses
Users may not:
- Rely on AI outputs as sole basis for regulatory filings
- Attempt to reverse engineer models
- Use outputs to train competing systems
- Misrepresent AI outputs as certified or verified results
9. Data Security & Processing Controls
Senseible applies access controls, environment separation, audit references, monitoring, and incident-response measures appropriate to the nature of the processing.
Enterprise data paths follow zero-trust principles: processing is restricted to explicitly authorized personnel and systems, with least-privilege and tenant-scoped access. This is an architectural approach, not a third-party certification.
10. Third-Party AI & Infrastructure
Senseible may use third-party AI tools or infrastructure components.
Senseible is not liable for third-party model behavior, outages, or policy changes.
11. Regulatory Alignment
Senseible aligns AI governance with:
- India Digital Personal Data Protection Act, 2023
- Emerging global AI governance principles
12. Limitation of AI Liability
Senseible disclaims liability for decisions made solely based on AI outputs.
AI systems are provided "as-is" and "as-available."
13. Policy Updates
This Policy may be updated to reflect regulatory, technical, or operational changes.
Continued use of the Services constitutes acceptance.
14. Contact
For questions, rights requests, or opt-out requests, contact the Data Officer:
Sneha Ghosh: shown on the live page at senseible.earth
Minarch Tower, Gurugram, 122003, Haryana, India
We aim to resolve requests within five to seven (5–7) business days, subject to applicable law and any identity-verification or request-complexity requirements.