AI Use & Disclaimer

1. Purpose of This Disclaimer

Pericls Ltd ("Pericls", "we", "us", or "our") operates a regulatory intelligence platform that uses artificial intelligence ("AI"), including large language models ("LLMs"), to help you work with complex, unstructured regulatory material.

This disclaimer explains, in plain terms, how we use AI, what its limitations are, and what we ask of you when you rely on AI-assisted output. It supplements the AI-related provisions of our Terms of Service and Privacy Policy; where there is any conflict, the Terms of Service prevail.

We ask you to read and acknowledge this disclaimer before using AI-assisted features so that you understand — and accept — the nature and limits of the technology.

---

2. How Pericls Uses AI

We use AI and automated processing to power core features of the platform, including:

• Regulatory analysis and summarisation of laws, rules, and guidance
• Obligation extraction — identifying discrete requirements within regulatory text
• Compliance gap analysis — comparing your stated posture against applicable obligations
• Regulatory roadmaps and horizon scanning — surfacing and prioritising changes
• Document analysis — reading and interpreting documents you upload
• Report and content generation — drafting summaries, assessments, and narratives

To deliver these features, we use third-party LLM providers, currently OpenAI, Anthropic, and Google (Gemini), all based in the United States. How your data is handled when it is sent to these providers — including that your data is not used to train their foundation models — is described in our Privacy Policy and Sub-Processor List.

---

3. Limitations of AI — Please Read Carefully

You acknowledge and accept that AI and LLM technology is inherently probabilistic and imperfect. In particular:

• Hallucination: AI systems can generate output that is fabricated, invented, or unsupported — including plausible-looking but non-existent citations, obligations, dates, figures, or legal references. This can happen even when the output appears confident and authoritative.
• Inaccuracy and incompleteness: AI output may be factually wrong, may misinterpret source material, or may omit relevant requirements, exceptions, or nuances.
• Currency: AI models and their underlying data have knowledge cut-offs and may not reflect the most recent regulatory changes, amendments, or interpretations.
• Context and jurisdiction: AI may misapply rules across jurisdictions, conflate similar-but-distinct regimes, or fail to account for facts specific to your business.
• Non-determinism: The same input can produce different output on different occasions.
• No understanding or judgment: AI does not "understand" law or your business and does not exercise professional judgment. It predicts likely text; it does not verify truth.

Because of these limitations, AI-generated output may be incorrect, misleading, or unsuitable for your circumstances, and must not be relied upon on its own.

---

4. Not Legal or Professional Advice

Pericls is a software tool, not a law firm, and does not provide legal, regulatory, financial, or other professional advice. AI-assisted output — including any analysis, roadmap, gap assessment, or generated report — is provided for informational and administrative purposes only and does not create any professional or advisory relationship.

You remain solely and fully responsible for your own regulatory compliance and for obtaining independent, qualified professional advice where appropriate.

---

5. Human-in-the-Loop — Your Responsibilities

Pericls is designed to be used on a human-in-the-loop basis: AI assists a qualified human who reviews, verifies, and decides. You agree that you will:

• Independently verify all AI-assisted output against authoritative primary sources before relying on or acting upon it;
• Not treat any AI-generated output as a final determination of your legal or compliance obligations;
• Apply your own professional judgment and seek qualified advice for material decisions; and
• Not use AI-assisted features for any purpose where an undetected error could cause harm without independent human review.

---

6. Automated Decision-Making

Pericls does not use AI to make decisions that produce legal or similarly significant effects about you or any individual without human involvement. AI features provide analytical assistance to support human decision-making. For more on your rights in relation to automated processing under data protection law, see our Privacy Policy.

---

7. No Warranty for AI Output

To the maximum extent permitted by law, AI-assisted output is provided "as is" and without warranty of any kind, including as to accuracy, completeness, currency, or fitness for a particular purpose. The limitations and exclusions of liability set out in our Terms of Service apply to AI-assisted output.

---

8. Changes to This Disclaimer

We may update this disclaimer to reflect changes in the AI technologies, providers, or practices we use. When we make material changes, we will update the version and effective date above and, where the disclaimer requires acknowledgement, ask you to review and acknowledge the updated version.

---

9. Acknowledgement

By acknowledging this disclaimer, you confirm that you understand how Pericls uses AI, that AI-assisted output can be inaccurate or fabricated, and that you are responsible for independently verifying such output before relying on it.

---

10. Contact

General Enquiries: Email: legal@pericls.com

Customer Support: Email: support@pericls.com

Postal Address: Pericls Ltd Charterhouse Millburgh Hall, Graffham, Petworth England, GU28 0QH

---

Document History

---

This AI Use & Disclaimer is provided in English. If there is any conflict between a translated version and the English version, the English version shall prevail.