How 10spring Uses AI

We build marketing and revenue systems for all types of companies. We use AI in our own work, and we think you are entitled to know exactly how we use AI.

This policy covers our own use of AI. It's not a description of the AI systems our clients build.

1. Where We Use AI

We use AI systems in the following parts of our work:

• Content production: Drafting, outlining, editing, and optimizing written material.

• Data classification: Categorizing free-text responses, such as "how did you hear about us" answers, at scale.

• Research and analysis: Summarizing sources, structuring findings, drafting analysis for human review.

• Measurement: Querying AI assistants to observe how they answer questions in a client's category, and how they cite sources.

• Internal tooling: Building and running internal software that assembles reports and dashboards from connected data sources.

• Operations: Transcription, meeting summaries, and routine document drafting.

We do not use AI to make final decisions about strategy, spend allocation, or client recommendations. A named person owns each of those.

2. Your Data and AI Systems

What we send to AI providers. In the course of analysis and content production, we may process client-related material, including campaign performance data, website content, research inputs, and anonymized customer feedback, using AI systems operated by third-party providers. All data is sanitized prior to use so that your company 's identifying information is not used.

The terms we require. We use business or enterprise tiers with contractual terms that prohibit the provider from using our inputs or outputs to train their models, and we prefer providers offering zero or limited data retention.

 What we do not do:

• We do not paste client data into consumer or free-tier AI tools that reserve the right to train on inputs.

• We do not upload complete client CRM exports, contact databases, or customer lists to general-purpose AI systems.

• We do not process special categories of personal data (health, financial account, biometric, or government-identifier data) through AI systems. Our engagements do not require it.

• We do not permit AI providers to use client data to train their models.

Where a client's own agreements or policies restrict AI processing further, those restrictions govern. Please notify us during your onboarding, and we will configure the engagement accordingly, including running specified work without AI assistance.

3. Human Accountability

Every deliverable has a named human owner who is accountable for its accuracy, quality, and appropriateness. AI assistance does not transfer that accountability. Specifically:

• No AI-assisted work product is delivered to a client without review by the person accountable for it.

• Factual claims, statistics, and citations in published material are verified against primary sources before publication. AI systems generate plausible-sounding errors, and we treat every unverified claim as unverified.

• Strategic recommendations, budget guidance, and measurement interpretations are authored by our team, not generated.

• Where work is technical or high-stakes, it receives additional review by a second team member.

4. Accuracy and Verification

AI systems produce fluent output that can be wrong. Our standard practice:

• Primary sources are cited for factual and statistical claims in published research;

• Data reported to clients comes from measured systems: analytics platforms, CRM, search consoles, citation-tracking tools. Not from AI-generated estimates;

• Where a number is an estimate, a model output, or a sample, we label it as such.

• We do not present AI-generated projections as measured results.

5. Disclosure of AI-Assisted Content

We use AI assistance in producing written material, as described in this policy. We do not label individual articles, because our standard is the same regardless of the tools used: a named human is accountable for every published piece, and factual claims are verified against primary sources. Where a client requires per-asset disclosure or prohibits AI assistance for specific deliverables, we follow the client's standard.

6. What We Will Not Do

• We will not use AI to generate false, misleading, or deceptive content, or to fabricate data, quotes, testimonials, reviews, or case study results.

• We will not use AI to impersonate real people, or to generate synthetic media depicting real individuals, without their explicit permission.

• We will not use AI to produce content designed to manipulate search or AI systems through deception, including fabricated expertise, fake authorship, and manufactured citations.

• We will not use AI to generate high-volume, low-value content for gaming rankings or citations.

• We will not use AI to make or automate decisions about individuals that produce legal or similarly significant effects.

• We will not use AI systems, enrichment tools, or automation to conduct cold outreach on our own behalf or on behalf of clients.

7. Security and Access

• We access client systems using individually credentialed, least-privilege, and, where possible, read-only access.

• Internal tools we build connect to client systems through scoped credentials and retain the minimum data required to operate.

• We do not retain copies of client data beyond what our engagement requires, and we return or delete client data on termination as set out in our services agreement.

• Access is reviewed periodically and revoked promptly when personnel or engagements change.

8. Our Principles

We hold a few positions about AI in this work:

• AI changes production, not accountability. Tools change constantly. The obligation to be right does not.

• Speed is not the point; correctness is. AI makes it cheap to produce more marketing material than anyone wants to read. We use it to work faster on things worth doing, not to flood channels,

• Measurement should get more honest, not less. AI has made marketing claims easier to generate and harder to verify. We publish what our measurement can and cannot establish, and we do not guarantee outcomes that no one can guarantee — including rankings and citations in AI systems.

• Human judgment is the scarce input. Strategy, taste, and the decision about what matters are the parts clients pay for. Those stay with people.

• We share what we learn. We publish our methods, our measurement standard, and our research, including the parts that did not work.

Questions? Email: legal@10spring.io

Last updated: 08/01/2026