The MCS Agency

Effective Date: July 18, 2025

AI Policy

This page explains how The MCS Agency uses AI in client work. It covers our standards for ethics, accuracy, privacy, and human oversight. It applies to our staff, contractors, and partners.

Principles

• People first. Strategy, judgment, and final decisions stay with humans.
• Transparency. We disclose meaningful use of AI in our process.
• Accuracy. Humans verify facts, sources, and context before delivery.
• Equity. We check for bias and write in inclusive, plain language.
• Privacy. We minimize data sharing and protect client information.
• Accountability. We document choices, keep records, and fix mistakes.

Definitions

AI tools. Software that generates or analyzes text, images, audio, or data.

Generative AI. Tools that create new content from prompts or inputs.

Client data. Any information a client shares with us, including text, files, visuals, or access to systems.

How We Use AI

Content Support

Brainstorms, outlines, first drafts, summaries, and tone adjustments. Every deliverable is written or fully edited by our team to match the client voice.

Research Assistance

Source mapping, pattern finding, and synthesis. Humans verify facts and add citations where needed.

Creative Exploration

Concept boards, headline options, design variations, and asset tagging. Designers and strategists make final selections.

Workflow Efficiency

Transcription, meeting notes, task routing, and quality checks — so our team can focus on higher-value work.

Analytics Support

Aggregating campaign data for human analysis and insight.

What We Do Not Do

• No unreviewed AI output in client deliverables.
• No training of external models on client data without explicit written consent.
• No deepfakes, synthetic testimonials, or fabricated quotes.
• No automated decisions that affect people without human review.
• No use of AI to imitate a person's voice or likeness without clear permission.

Data and Privacy

Data Minimization

We share the least amount of client data needed to complete a task.

Redaction

We remove names, personal details, and sensitive fields before using AI tools, unless a client directs us otherwise in writing.

Storage

We prefer tools that allow data control and limit retention. We adjust settings to disable training on our prompts and outputs when possible.

Access Control

Only assigned team members can view client data. We use strong authentication and role-based permissions.

Retention

We keep working files only as long as needed for the project, or as required by contract or law.

Vendor Standards

• We select vendors that publish security practices and offer robust controls — including encryption at rest and in transit — and that undergo independent security audits such as SOC 2 or ISO 27001.
• We review vendor policies for data retention, training, and deletion paths.
• We use private or enterprise features when available to add protections for prompts and outputs.

Human Review and Quality

• A strategist or editor reviews every deliverable for accuracy, clarity, tone, voice, and brand alignment.
• Facts are checked against primary or credible sources. Citations are added when appropriate.
• Images and design assets are reviewed for rights, permissions, and accessibility.
• We test content for plain language, readability, and inclusive framing.

Bias and Accessibility

• We assess prompts and outputs for harmful bias and stereotypes.
• We use inclusive style guides and reference materials.
• We design with accessibility in mind, including alt text, contrast, captions, and logical content structure.

Intellectual Property

• Clients own final deliverables as set in our contract.
• We license stock and fonts properly and keep proof of license.
• We do not include materials we do not have the right to use.
• When AI-assisted tools are used to create or edit an asset, we confirm the license terms allow client use.

Disclosures and Labels

We disclose meaningful AI assistance in our process. Language you may see on deliverables:

Content label. This content was created with assistance from AI tools and finalized by The MCS Agency team.

Image label. This image was created or edited using AI and finalized by The MCS Agency team.

Client Choices

• You can request that we avoid AI assistance in your work. We will confirm scope, timelines, and cost based on that choice.
• You can approve or decline the use of specific tools in writing.
• You can request a record of which tools supported a deliverable.

Incident Response

If we discover an accuracy issue, bias concern, or privacy exposure linked to AI use, we will notify the client promptly, investigate, correct the issue, and document the fix.

We review root causes and adjust prompts, tools, or processes to prevent repeat issues.

Governance and Training

• Staff and contractors receive guidance on prompt design, data handling, bias checks, accessibility, and disclosure.
• Leads review new tools before use in client work.
• We conduct periodic audits of projects that used AI assistance.

Scope and Exceptions

This policy applies to all client projects and internal work that may affect clients. Any exception must be approved in writing by agency leadership and documented in the project file.

Last updated: June 17, 2026