BlackSwan Foresight is in early deployment. Where a safeguard is designed but not yet built, this page says so rather than implying otherwise.
Your data stays yours
- We do not use client data to train, fine-tune or improve any public or shared AI model.
- Forecasts from your ledgers are fitted on your data only and are not pooled with other clients' data.
- The credit-risk model used by Revenue and Customer Foresight is a single shared model. It is currently trained on synthetic data, not on any client's receivables, and any future retraining policy will be published in the AI Model Card before it changes.
Separation between clients
Each client's financial and customer data is kept separate by design and used only to produce that client's readouts. Production multi-client hosting with per-client authentication is designed and being built. Until it is live, each client engagement runs separately, and the specific environment and controls for your data are set out in your engagement agreement.
Storage, encryption and access
- This website is served over HTTPS, so information you send through it, including the assessment, is encrypted in transit.
- Client data is accessed only by the BlackSwan team members working on your engagement.
- Where your data is stored, how it is encrypted and how long it is kept are agreed in writing before you share any data with us.
- Client data is deleted at the end of an engagement or earlier on request, unless the law requires us to keep it.
AI providers
Where the system uses a third-party AI model, it does so through standard inference calls that are not used to train the provider's models. Our current agent layer uses Anthropic's Claude.
People stay in charge
Every output is a forecast or a recommendation. The system never makes payments, changes your books or takes financial decisions on your behalf.
Governance
Our design and data handling are guided by ISO/IEC 42001 and the NIST AI Risk Management Framework. We use them as design guidance and do not claim certification against either. Our founder holds an AI Security & Governance certification from Securiti. Methods, data sources and known limitations are disclosed in our AI Model Card.
Report a concern
If you believe you have found a security issue, or have a question about how your data is handled, email shankar.dv@blackswanconsulting.in. We treat security reports as a priority and will acknowledge yours as quickly as we can.