AI Policy
Austley's AI Policy
Our Stance
We help marketing and communications teams put AI to work. That means we use AI every day, openly, and with discipline.
AI is a tool. It is not a replacement for the strategy, judgment, and craft that define great communication. Every Austley deliverable is shaped by experienced humans. AI helps us move faster, think more broadly, and serve clients better, but it does not have the final word.
This policy exists so our clients know exactly how we use AI, what we will not do with it, and how we protect the trust they place in us.
Who This Applies To
This policy applies to every Austley team member and to every freelancer, contractor, vendor, and partner who works on Austley client engagements. We require contractors to follow these standards.
Definitions
AI: Software that performs tasks typically requiring human intelligence, including generating language, images, audio, and analysis.
Generative AI: AI that creates new content such as text, images, video, audio, and code.
Large Language Models (LLMs): Generative AI trained on large text datasets that produce human-like writing. ChatGPT, Claude, Gemini, and Copilot are examples.
Private AI: AI tools configured so that client inputs are not used to train public models, are not shared with other users, and are retained only as needed to deliver the service. Austley uses private AI configurations wherever possible.
Public AI: Free or consumer-tier AI tools where inputs may be retained, used to train models, or accessed by the provider for service improvement. Austley does not put client information into public AI tools.
Our Principles
1. Human always. AI supports our work. People do the work. Every client deliverable is reviewed and approved by an experienced Austley team member.
2. Transparent. We tell clients how we use AI on their work. We disclose tools, methods, and any AI-generated material in deliverables on request.
3. Protected. Client information stays inside private, enterprise-tier tools. We do not feed client data to public AI models.
4. Responsible. We check our outputs for accuracy, bias, and originality before they reach a client or an audience.
5. Accountable. When AI assists our work, the responsibility for the result still belongs to Austley, so AI is neither blamed nor credited.
How We Use AI
Research and synthesis
We use AI to scan literature, summarize sources, find patterns across documents, and surface angles we might miss. We always verify facts and cite primary sources before they reach client materials.
Ideation and creative development
We use AI to expand the option set. Draft headlines, concept directions, message variants, and naming options. The selection, refinement, and final craft are human decisions.
Drafting and content production
We may use AI to produce first drafts of long-form content, social copy, presentations, and internal documents. Every published piece is rewritten, fact-checked, and brand-aligned by an Austley team member. We do not publish AI-generated content unedited.
Workflow and operations
We use AI to organize information, format content, build structured outputs, transcribe meetings with consent, and accelerate repeatable internal tasks.
Strategy and analysis
We use AI to synthesize large amounts of qualitative or quantitative input. The recommendations and decisions are made by senior Austley strategists.
Tools we use
We work primarily in enterprise and paid tiers configured for privacy. Our active stack includes ChatGPT Business, Claude Max, Google Gemini Ultra, Perplexity Pro, and Microsoft Copilot. We review this list regularly. Clients can request the current configuration on request.
What We Will Not Do
We do not publish content that is 100% AI-generated. Every published piece undergoes human review.
We do not put confidential, sensitive, or client-proprietary information into public AI tools.
We do not allow client data to be used to train public AI models.
We do not use AI to create deepfakes, voice clones, or synthetic likenesses of real people.
We do not use AI to alter the appearance, ethnicity, or expression of people in client photography or video.
We do not fabricate quotes, testimonials, sources, or research using AI.
Human Review
Every deliverable that involves AI is reviewed by an Austley team member before it leaves our hands. That review checks for:
Factual accuracy, including verification of statistics, sources, names, and quotes
Brand voice and tone alignment with the client's guidelines
Bias, stereotyping, and assumptions that should be questioned
Originality, including a check against the source material AI may have drawn from
Strategic fit with the client's broader goals
If something cannot be verified, we cut it or replace it with something we can verify.
Disclosure to Clients
We are open about how we use AI. If you want to know how AI factored into a specific piece of work, ask. We will tell you what tools were used and where.
Clients with stricter standards, including regulated industries, can request our enterprise AI addendum for more detail.
Data Handling and Privacy
We handle client data in line with GDPR, CCPA, and any client-specific data requirements written into our agreements.
We use enterprise and paid AI tools configured to keep client inputs out of public training data. Where a vendor offers data residency, retention controls, or training opt-outs, we configure for the strictest setting available.
We do not share client information across engagements. We do not retain client data inside AI tools longer than the engagement requires.
For clients in regulated industries, including health technology, financial services, insurance, and legal services, we operate to a higher standard. We can provide detailed configuration documentation, our enterprise AI addendum, and a tool-by-tool data flow on request.
Sustainability
We recognize that AI tools carry an energy cost. We use them with intention to enhance the efficiency and quality of our work.
Governance and Accountability
This policy is owned by Austley's leadership and reviewed at a minimum once per year, or sooner if regulations, client needs, or our tools change materially.
Austley team members are trained on this policy at onboarding and at every annual update. Training covers safe use of AI tools, data handling, bias awareness, and verification practices.
Concerns, questions, and incidents related to our AI use can be raised at any time by emailing ai@austley.com. We take every report seriously and respond within five business days.
If something goes wrong with AI-assisted work, we own it. Then we take the necessary steps to fix it, update our practices if necessary, and learn from it, so it doesn't happen again.
The Future of Work
AI is changing fast and so is the regulation around it. We will continue to refine this policy, update our tools, and raise our standards. Clients can expect us to be ahead of these changes.
If you are a client, prospect, or partner with questions about how this applies to your work with us, feel free to contact us at any time.
