AI Features & Usage Statement

AI Features & Usage Statement

Loncom Consulting Ltd — Company Number: 13572610 — United Kingdom
Last updated: March 2026

1. Purpose

This document explains how Loncom Consulting Ltd uses artificial intelligence ("AI") tools and features in the delivery of our services, the safeguards we apply, and the benefits our clients can expect.

We see AI as an accelerator for analysis, improvement and application. It is never a replacement for expert human judgement.

2. Scope

This statement covers our use of the following tools:

  • Cursor — AI-assisted development environment for code and automation
  • Claude — large language model for reasoning, drafting and analysis
  • ChatGPT — large language model for research, drafting, code support and explanation
  • HubSpot AI features — including content assistants, AI insights and AI-powered CRM capabilities

It applies to work we perform for clients across CRM development and implementation, ecommerce and operational data analytics, reporting, dashboards and forecasting, and integrations, data migration and data quality projects.

3. Our AI Principles

We use AI in line with four core principles:

  1. Human-in-the-loop
    AI supports our consultants. It does not make final decisions. All AI-generated outputs that affect client environments are reviewed, validated and approved by a human expert before being applied.
  2. Data protection and security by design
    We minimise the personal or sensitive data sent to AI tools. We do not knowingly send full production datasets, credentials or highly confidential information into generic AI models. Where client data must be referenced, we apply redaction, sampling or synthetic examples wherever possible.
  3. Transparency
    We are open about where and how AI is used in your project. Where we embed AI features in your stack (for example HubSpot AI), this is documented in the solution design.
  4. Quality and accountability
    AI is used to improve speed, coverage and consistency, not to lower standards. Loncom remains accountable for the quality and correctness of all deliverables, regardless of which tools are used to produce them.

4. How We Use AI Tools

4.1 Cursor (AI-Assisted Development)

We use Cursor as an AI-powered development environment to accelerate the writing and refactoring of code for integrations, data pipelines and automation scripts, to generate test scaffolding and validation checks, and to identify potential improvements in existing codebases.

Safeguards: We avoid pasting secrets, API keys or full production tables into prompts. Any code generated or refactored via Cursor is reviewed, tested and version-controlled in exactly the same way as manually written code.

4.2 Claude and ChatGPT (Reasoning, Drafting and Analysis)

We use Claude and ChatGPT to assist with drafting documentation, runbooks and technical specs, summarising complex requirements and data dictionaries, suggesting data models and workflow patterns, and providing code snippets that our team then reviews and adapts.

Safeguards: We focus on structures and patterns rather than bulk transfer of raw personal data. Where client data must be referenced, we prefer sampling, masking or synthetic data. All recommendations are verified against your actual business rules, compliance requirements and technical constraints.

4.3 HubSpot AI Features

When HubSpot AI features are part of the agreed design, we may configure and enable the Content Assistant for drafting emails, landing page copy and CTAs, AI-assisted property insights to support data quality and segmentation, and AI tools in workflows and reporting to augment lead scoring, routing or summarisation.

Safeguards: Any AI-generated content in HubSpot is reviewed and edited by a Loncom consultant or by the client before use. We configure AI features in line with your compliance stance and ensure that all AI within HubSpot is consistent with your brand, tone of voice and privacy commitments.

5. Typical Client Benefits

By using AI responsibly, clients benefit from:

  • Faster delivery — quicker drafting of documentation, specifications and code scaffolding
  • Higher coverage — more scenarios and edge cases considered during design and testing
  • Clearer documentation — better-structured guides, playbooks and training materials
  • More iterative experimentation — the ability to test more options for scoring models, workflows and dashboards before committing

The result is not AI for its own sake, but more robust CRM, data and integration solutions delivered faster and with greater clarity.

6. Data Handling and Privacy

Our AI usage sits within the broader controls defined in Loncom's Privacy Policy. In practice this means:

  • We treat any client data used in AI prompts as confidential and only use it where necessary for the task
  • For processor work, we follow the client's data processing agreement and instructions
  • We apply least-privilege access to the systems and datasets we connect to, regardless of whether AI is involved
  • Where an AI vendor offers specific data residency or no-training options under enterprise terms, we factor that into our technology selection and configuration

If you have stricter requirements, for example a mandate that no production personal data is ever sent to external AI services, we can design a delivery approach that fully respects this.

7. Human Oversight and Quality Control

For every use of AI in a client project:

  • A named Loncom consultant remains responsible for the output
  • AI-generated code is subject to the same review, testing and deployment processes as any other code
  • AI-generated documentation or content is reviewed for accuracy, clarity and alignment with your internal language and policies
  • We do not enable self-changing production workflows based solely on AI decisions

8. Client Choices and Opt-Out

We recognise that organisations have different comfort levels with AI. We therefore offer the following options:

  • No external AI for your data — we avoid using Claude or ChatGPT with any identifiable client data and rely only on internal tools and established patterns
  • HubSpot-only AI — we use only the AI features directly embedded in your HubSpot instance, within your own governance framework
  • Full AI-enhanced delivery — we use the full toolkit including Cursor, Claude, ChatGPT and HubSpot AI with our standard safeguards for maximum speed and depth

Your preferences can be captured in the Statement of Work or Data Processing Agreement and reflected in our delivery approach from the outset.

9. Governance and Continuous Improvement

AI capabilities evolve quickly. We periodically review the AI tools we use against security, privacy and reliability criteria, update our internal guidelines and training to ensure responsible and effective use, and adjust our approach as regulatory expectations, client requirements and vendor capabilities change.

Any material change in how we use AI that affects your data or your solution design will be communicated to you and, where relevant, reflected in updated project documentation.

10. Contact Us

If you have any questions about our use of AI tools or would like to discuss your specific requirements, please contact us:

Email: info@loncomconsulting.com
Company: Loncom Consulting Ltd, Company Number 13572610, United Kingdom