AI Transparency Statement

1. What AI We Use

ClearSafe Ltd uses Claude, a large language model developed by Anthropic PBC, accessed via the Anthropic API. Claude powers both the Felix AI advisor and all document generation features on the platform.

2. What AI Does on This Platform

AI is used to power the following features:

• Felix — conversational H&S compliance advisor
• Risk assessment generation (general, COSHH, manual handling, working at height, DSE, lone working, machinery & equipment)
• Fire risk assessment generation (standard and HMO)
• Method statement (RAMS) generation
• Health & Safety policy generation
• Hazard identification guidance
• Photo analysis for hazard identification
• Document upload and analysis (scoring and gap identification)
• RIDDOR reporting check
• Incident severity assessment

3. What AI Does Not Do

AI on this platform does not:

• Make final compliance decisions on your behalf
• Replace the review of a competent person as defined under the Management of Health and Safety at Work Regulations 1999
• Provide legally binding advice
• Access real-time legislation updates — the AI model has a knowledge cutoff date and may not reflect the most recent legislative changes

4. How We Label AI Content

ClearSafe Ltd fulfils its transparency obligations through the following measures:

• All AI-generated PDF exports include an amber disclaimer clearly identifying the content as AI-assisted
• An export acknowledgement modal requires users to confirm they have reviewed AI-generated content before downloading
• Felix is clearly identified as an AI system in the chat interface
• PII warnings are displayed on all upload pages where data is sent to the AI provider

5. EU AI Act Classification

Under the EU AI Act (Regulation (EU) 2024/1689), ClearSafe Ltd's use of AI is classified as limited risk. This classification requires transparency obligations, which are met through the labelling and disclosure measures described in Section 4 above.

ClearSafe Ltd does not use AI for any purpose classified as high risk under Annex III of the EU AI Act.

ClearSafe Ltd is committed to responsible AI development in line with the UK Government's voluntary AI Code of Practice and the DSIT AI Principles. We are committed to transparency, accountability, and human oversight of all AI outputs.

6. UK GDPR and AI

ClearSafe Ltd does not make automated decisions that affect users' legal position. All AI outputs are draft documents presented for user review. No document is adopted or given legal effect until the user reviews, modifies (if necessary), and signs it off.

The lawful basis for AI processing is legitimate interests (providing the contracted service) and contractual necessity.

For full details on how we process personal data, see our Privacy Policy.

7. Data Sent to AI

When you use AI features, the following types of data may be included in prompts sent to Anthropic:

• Company name, industry, and site/premises name
• Hazard descriptions, control measures, and risk ratings
• Incident descriptions and injury details
• Uploaded document content (text extracted from PDFs, DOCX, XLSX files)
• Uploaded photos (for hazard identification)
• Messages sent to the Felix chat advisor

Anthropic processes this data under a data processing agreement. Anthropic does not use data submitted via the API for model training. For full sub-processor details, see our Privacy Policy.

8. Your Rights

Users retain full control over all AI outputs and are not bound by them until sign-off. You may edit, reject, or delete any AI-generated content at any time before adopting it as a compliance document.

Users who require professional H&S consultancy support beyond the platform's AI-assisted tools may contact enterprise@getclearsafe.com to enquire about consultancy services.

9. Legislative Currency

The AI model powering ClearSafe Ltd has a knowledge cutoff date and may not reflect the most recent legislative changes, HSE guidance updates, or case law developments.

Users are responsible for independently verifying that AI-generated content reflects current legislation applicable to their workplace. ClearSafe Ltd accepts no liability for content that does not account for legislative changes made after the AI model's knowledge cutoff.

10. Contact

If you have questions about ClearSafe Ltd's use of AI, contact us:

ClearSafe Ltd

Company number: 17093354

66 Paul Street, London, England, EC2A 4NA

Phone: +44 (0)330 133 0620

Email: hello@getclearsafe.com