AI Transparency Policy
Last updated: September 26, 2026
The short version
No smoke. No mirrors. I use artificial intelligence in this business. AI helps me research, think, draft, organize, analyze, design, automate, troubleshoot, and build faster. It does not replace my judgment, lived experience, creative direction, accountability, or relationship with my clients. Samantha Diaz DBA The Hot Mess Collective remains responsible for the work carrying The Hot Mess Collective name.
How I use AI
Depending on the project, AI may assist with:
Research, brainstorming, and outlining.
Drafting, editing, and refining written content.
Workflow mapping and systems design.
Coding, automation, troubleshooting, testing, and quality assurance.
Organizing, comparing, and analyzing information.
Developing visual concepts, imagery, and production direction.
Summarizing material and exploring alternative approaches.
Documentation and internal operational work.
Supporting repetitive or administrative parts of business processes.
The amount and type of assistance varies by project. Some work uses AI heavily behind the scenes. Some uses it minimally. Some does not require it at all.
What stays human
AI does not make final strategic decisions for The Hot Mess Collective or its clients. I determine the direction, apply context, protect the intended voice, challenge weak recommendations, make judgment calls, and decide what belongs in the final work.
Material public content and client deliverables receive human oversight appropriate to how they will be used and the consequences of getting them wrong. I do not present an AI system as a licensed professional or treat its output as a substitute for legal, financial, accounting, tax, medical, mental-health, regulatory, compliance, or other licensed professional advice.
Client work
Unless we agree otherwise in writing, AI-assisted tools may be used in portions of client work. That does not mean a raw AI response is passed off as a finished deliverable.
Client needs, context, strategy, brand standards, approvals, existing systems, and project agreements remain the foundation of the work. The applicable service agreement or statement of work controls ownership, confidentiality, licensing, approval, and delivery terms for each engagement.
If a project requires an AI-free workflow, prohibits particular providers, or requires specific restrictions on how information may be processed, tell me before the work begins and before confidential information is shared. I will confirm whether those restrictions are feasible and whether they affect the scope, timeline, tools, or price of the engagement.
AI providers and third-party tools
THMC may use AI capabilities provided directly by AI companies or incorporated into other software, platforms, design tools, automation systems, productivity tools, or business applications. Those providers control their own infrastructure, security practices, retention practices, model behavior, availability, policies, and terms.
THMC does not control a third-party AI provider merely because the provider is used as part of a workflow. Where appropriate, I consider the sensitivity of the information, the purpose of the task, and the available provider controls before using AI with client material.
Confidentiality and sensitive information
I do not intentionally enter passwords, complete payment-card information, government identification numbers, medical records, or unnecessary sensitive personal or confidential information into general-purpose AI tools. When AI is used with client-provided material, I aim to limit the information to what is reasonably necessary for the task and remove or reduce identifying information when practical.
Clients should not send unnecessary sensitive information through website forms, ordinary email, or project systems. No online platform is risk-free. Any special confidentiality, privacy, security, regulatory, contractual, or data-handling requirement must be identified and agreed to before the relevant information is shared or processed.
Accuracy, limitations, and bias
AI can be confidently wrong.
It can miss context, oversimplify a problem, misunderstand instructions, invent information, reproduce bias, rely on outdated information, or produce something that looks polished while being fundamentally incorrect.
I review AI-assisted work based on its purpose and level of risk, verify important factual claims against appropriate sources when necessary, and increase human review when the stakes are higher. Human review reduces risk. It does not make every error impossible.
Automation and AI-assisted systems
Some THMC systems may use AI or automation to route information, generate drafts, organize records, classify inputs, trigger follow-up actions, or assist with repetitive operational work. Automation is not the same thing as judgment.
THMC does not intentionally delegate legal, medical, financial, regulatory, or similarly high-impact decisions to an AI system merely because the technology is capable of generating an answer. Where a workflow affects a client relationship, purchase, application, or material business decision, the level of human involvement depends on the nature and consequences of that workflow.
Creative work and disclosure
AI may assist with writing, editing, visual concepts, imagery, production, and other creative work. Final creative direction and decisions remain human-led. I do not label every instance of routine AI assistance, such as brainstorming, editing, summarization, research support, background removal, layout exploration, or drafting.
I will disclose materially AI-generated or AI-altered media when failing to disclose it could reasonably mislead someone about a real person, event, endorsement, client result, source of evidence, or other material fact. THMC will not fabricate testimonials, client outcomes, credentials, reviews, endorsements, or proof. Where a platform or applicable law requires additional disclosure of AI-generated or altered material, THMC will follow those requirements.
No guarantee of AI availability
AI products and providers change. Features can disappear. Models behave differently after updates. Pricing changes. APIs break. Usage limits appear. Platforms shut down products entirely.
A THMC workflow that uses a third-party AI service may therefore require adjustment, replacement, or human intervention if that provider changes its technology or terms. THMC does not guarantee the permanent availability or performance of a third-party AI provider.
Changes to this policy
AI tools, platform practices, laws, and the way THMC uses these technologies will continue to evolve. This policy may be updated as those practices change. The current version and most recent update date will remain available on this page.
Questions or restrictions
Questions about this policy or the use of AI in a potential engagement may be sent to:
Samantha Diaz DBA The Hot Mess Collective
sam@thehotmesscollective.com
If you have a specific restriction, bring it up before sharing confidential information or beginning the work so we can determine whether the project is a fit.