Introduction:

Marketing Alchemist uses artificial intelligence tools to deliver marketing and event consultancy services, from drafting documents and generating research to cleaning and structuring B2B data. We use AI to enhance our capabilities while keeping human expertise, ethical standards, and client trust at the centre of everything we do. This policy sets out how AI is used, what data it is permitted to process, and the controls that apply.

​

1. AI as an Assistant, Not a Replacement:

AI is used at Marketing Alchemist to support the consultant's work, not to replace human judgement. Every AI-generated output, whether written content, code, or structured data, is reviewed manually before it is incorporated into any client deliverable. Human-in-the-loop review is applied on every project, without exception. AI is used for support tasks such as drafting, research, code generation, image or mockup generation, data cleaning, and quality assurance. Final deliverables are approved and refined by the consultant before they reach the client.

​

​

2. Approved AI Tools

Marketing Alchemist's primary AI tool is Claude (Anthropic), used on a paid Claude Max subscription. Under Anthropic's terms for paid plans, customer inputs and outputs are not used to train Anthropic's models by default. Other AI tools may be used for specific tasks (for example, image or video generation); where they are, they are selected on the basis of their privacy settings, data handling terms, and suitability for the task.

​

3. What Data Goes into AI Tools

AI use at Marketing Alchemist is restricted to:

• B2B business contact data (company name, business email, business phone, job title, publicly listed social media handles)

• Publicly available business information

• Non-confidential project content (drafts, briefs, structure, research summaries)

The following categories of data are never entered into an AI tool:

• Consumer personal data

• Special category data (race, political opinions, religious beliefs, trade union membership, health, sex life, sexual orientation, genetic data, biometric data)

• Financial data, payment information, or account credentials

• Confidential client information that the client has identified as restricted

• Personal data where the client has not authorised AI-assisted processing

Where a client imposes stricter restrictions than those set out here, the client's instructions apply.

​

4. Quality, Accuracy, and Fact-Checking

AI outputs are treated as working drafts, not as finished or authoritative results. All AI-generated written content is fact-checked, edited, and tailored to the client's voice and requirements. Generated code is read and tested against sample data before being run at scale. Data outputs (cleaned rows, enriched records) are spot-checked against source records to confirm fidelity. No AI output reaches the client without this verification step.

​

5. Ethical Use and Honesty of Framing

AI is used in line with Marketing Alchemist's commitment to honest, accurate, and respectful communication. Specifically:

• AI is not used to generate deceptive content, synthetic identities, deepfakes, or fabricated quotes

• AI-assisted outreach copy is reviewed to ensure recipients receive accurate information about who is contacting them and why

• AI is not used to make decisions that produce legal or similarly significant effects on individuals (for example, decisions relating to employment, credit, benefits, or legal status)

• Bias in AI outputs is mitigated through human review; the consultant's judgement shapes the final message and flags any output that is stereotyped, inaccurate, or inappropriate

​

6. AI Governance Framework

Marketing Alchemist operates within the following governance framework:

• Platform-level: Use of Claude is governed by Anthropic's published Usage Policy and Acceptable Use Policy

• Regulatory: AI use is aligned with the minimal-risk criteria set out in the EU AI Act and equivalent frameworks; prohibited use cases (for example, emotion inference from biometric data, or categorisation of individuals by special category data) are not undertaken

• Internal: Working rules restrict AI use to the data categories set out in Section 3 above; the consultant reviews AI use at the start of each new project to confirm it remains appropriate to the scope

• Client-level: Where a client has an AI governance framework or specific contractual requirements, Marketing Alchemist accepts and operates within those requirements

​

7. Data Retention and Chat History

AI chat history is managed on a per-project basis. On completion of a client project, chat sessions relating to that project are cleared, so no long-term AI-held record of client data is retained. The authoritative record of project work is held in Marketing Alchemist's working systems and delivered to the client; AI interactions are working sessions only.

​

8. Sub-processors and AI Use

Where Marketing Alchemist engages a sub-processor (for example, a research or telecalling agency), the sub-processor is bound by a written Data Processing Agreement that applies the same restrictions on AI use, confidentiality, and data handling set out in this policy and in the client's underlying agreement.

​

9. Intellectual Property and Legal Compliance

Marketing Alchemist respects intellectual property rights in all AI-generated and human-generated content. Where AI is used to generate imagery or graphics, outputs are reviewed to confirm they are original or suitably licensed for use. AI-assisted outputs produced as part of a client deliverable are owned by the client in accordance with the intellectual property provisions of the client agreement. We stay informed of evolving AI laws and guidelines (both in the UAE and globally) and adjust our practices as needed to remain compliant.

​

10. Transparency and Client Collaboration

We are transparent about our use of AI. Clients can ask how a piece of content or strategy was developed, and we will openly share the role (if any) that AI played. AI is used to serve client needs, never as a shortcut around quality, accuracy, or creative judgement. Where a client requests disclosure of AI use in a vendor assessment or procurement process, we will disclose in full.

​

11. Continuous Review

The AI landscape is evolving, and so is this policy. We review and update it at least annually, and more often if needed, to reflect changes in AI tools, regulatory requirements, and best practice. The current version is always available on our website with the effective date.​

​

For more information: contact@marketing-alchemist.co