Ai Policy

Effective date: 23 September 2025

Scope: This policy explains how we use Artificial Intelligence (AI) on our website, in client work, and internally.

1) Principles

·       Human oversight: People remain accountable for all final outputs.

·       Transparency: We disclose meaningful AI assistance where relevant.

·       Fairness: We check for bias and avoid harmful stereotypes.

·       Privacy-first: We minimise data, avoid sensitive inputs, and use safeguarded tools.

·       Lawful use: We follow GDPR and prepare for the EU AI Act.

2) What we use AI for

·       Drafting and refining copy, metadata, and ideas.

·       Summarising research; quality checks (clarity, spelling, tone).

·       Light data analysis on non-personal datasets.

·       Concepting visuals with licensed/permissioned assets.

We do not use AI to impersonate people, fabricate testimonials, or mislead.

3) Human review

All deliverables are reviewed by qualified staff for accuracy, legality, ethics, brand fit, and potential bias before publication.

4) Data protection (GDPR)

·       Role & basis: We act as controller or processor per contract and rely on a valid legal basis.

·       Minimisation: Only necessary data is entered; we prefer tools that do not train on inputs by default.

·       Special-category data: Not used in AI tools unless expressly instructed with safeguards and a DPA.

·       Your rights: Contact us to access, delete, correct, restrict, object, or request portability.

5) EU AI Act readiness

We monitor the Act’s staged application and will adjust transparency, governance and (where relevant) GPAI duties. Key dates include 2 Feb 2025 (prohibitions & AI literacy), 2 Aug 2025 (GPAI governance), 2 Aug 2026 (most rules), and further high‑risk provisions in 2027.

6) Third‑party AI tools

·       We use reputable providers and review input handling, security, model limits and sub‑processors.

·       We keep an internal register and prohibit unvetted tools.

7) Content labelling & IP

·       Where material is substantially AI‑assisted, we may label it (e.g., “AI‑assisted”).

·       We respect copyright and only use assets we own or are licensed for.·      

8) Quality, safety & limitations

AI outputs can be wrong or outdated. We mitigate with human review, source checks, and testing, and we do not rely on AI for regulated advice.

9) Your choices

Clients may opt out of AI‑assisted workflows. This may affect timelines or costs; we will discuss options.

10) Changes

We may update this policy as regulations and best practice evolve. Material changes will be posted with a new effective date. See our Privacy Notice for data specifics.

11) Contact

Optimise Marketing
Wexford Enterprise Centre
Strandfield Business Park, Kerlogue
Rosslare Road, Wexford
Ireland

Tel:+353 (0)53 9006063
office@optimisemarketing.ie