How MarginGlow uses AI.

1. What our AI does.

MarginGlow uses AI to generate business suggestions, summaries, draft answers, sustainability report content, task ideas and practical recommendations. These outputs are intended to support small business owners, not replace their judgement.

2. AI-generated content.

Content generated by MarginGlow may be produced or assisted by AI. Users should review AI-generated text before using it publicly, submitting it in tenders, sending it to customers, changing advertising activity or making business decisions.

3. No professional advice.

MarginGlow does not provide legal, tax, accounting, financial, employment, insurance or regulated sustainability assurance advice. It may help organise information and draft useful content, but professional judgement may still be needed.

4. Business data and uploads.

Users may provide business information, documents, screenshots, invoices, receipts, statements, tender questions or sustainability data. MarginGlow uses this information to produce outputs for the user’s private workspace and to improve recommendations using anonymised operational patterns where appropriate.

5. Accuracy and limitations.

AI outputs can be incomplete, outdated or wrong. MarginGlow aims to provide clear, practical and useful recommendations, but users should verify important facts, calculations, claims and customer-facing statements before relying on them.

6. AI emissions transparency.

MarginGlow aims to track the estimated carbon footprint of AI tasks run in a user’s workspace and to help businesses reduce avoidable AI use by making previous answers, files, tasks and reports searchable. Better memory should reduce repeated questions, wasted Task Credits and unnecessary AI emissions.

7. Human control.

Users remain in control. MarginGlow suggestions should be treated as recommendations to review, edit, approve, ignore or improve. Where a workflow affects customers, advertising spend, sustainability claims, legal commitments or financial decisions, users should apply appropriate human review.

8. EU AI Act readiness.

MarginGlow is being designed with transparency, user control, clear AI disclosure, data minimisation, human review and responsible AI use in mind. This policy will be updated as EU AI Act guidance and implementation practice develops.

9. AI-generated imagery.

Our landing-page clone illustration is AI-generated. It depicts a fictional scenario and is not a photograph of a customer or a customer testimonial. We label this illustration next to the image so visitors can recognise its origin.

10. MarginGlow Signal and AI agents.

MarginGlow AI Signal allows authenticated external AI agents to request business opportunity analysis through Model Context Protocol (MCP). An agent supplies business details such as a name, public website and country. MarginGlow returns AI-generated analysis with supporting evidence, ranked opportunities and confidence information. These outputs can contain errors and should be reviewed before use.

Signal analysis uses public business information; this interface does not grant access to a customer’s private MarginGlow workspace or connected accounts. An analysis consumes a complimentary or paid Signal credit. Agents can check their Signal balance and request a Stripe payment link to buy an additional Signal. Users should set spending permissions with their calling agent and review its privacy and data-handling terms. Signal credits are separate from workspace task credits.

The calling application controls how it presents the analysis and any subsequent actions. MCP is a connection protocol, not a guarantee of accuracy, user consent or regulatory compliance. Carbon-footprint measurement and verified carbon-removal purchasing through this interface remain a development vision, not a capability promised by the current Signal analysis service.

11. Contact.

Questions about MarginGlow AI transparency, data use or responsible AI practices can be sent to support@taoclimate.com.