AI & LLMs Policy
We use artificial intelligence (AI) responsibly to draft, summarise and speed up internal work. We do not rely on AI as the final authority for engineering decisions.
When we use AI
- Drafting emails, reports and checklists that a human reviews before sending.
- Creating proposals, risk registers or method‑statement outlines that are then validated by an engineer.
- Generating code/scripts to automate repetitive tasks, reviewed before use.
- Early idea generation (options, pros/cons) — not a replacement for design calculations or approvals.
What we never use AI for
- Approving design or safety‑critical decisions.
- Replacing qualified sign‑offs or competent checks under CDM/BS 5975.
- Sharing confidential client data without consent.
Data handling
- We avoid sending personal or confidential information to AI tools unless a client asks us to and data is minimised/anonymised.
- Third‑party AI providers may process prompts to deliver the service; we do not allow them to use client data to train public models.
- Outputs are checked for accuracy and safety. If sensitive, they are reviewed by a senior engineer/manager before release.
- Contact info@eblom.co.uk to request removal of AI‑related content that includes your data.
Training opt‑out for clients
If you prefer your data not to be used in any AI assistance, email info@eblom.co.uk with subject AI OPT‑OUT. We will confirm and record the instruction for your projects.
LLMs.txt (plain‑text disclosure)
You can also publish a simple text file at the site root so visitors know our policy quickly. Suggested contents:
# LLMs.txt — AI disclosure for Eblom
We use AI/LLMs to assist with drafting and automation. Humans review all external deliverables.
We do not upload confidential information without consent, and we do not allow training on client data.
Contact: info@eblom.co.uk
Last updated: 2025-08-30