May 5, 2026 · 3 min read
How We Use AI — And Where a Human Always Steps In
AI is central to how Sublime Plus works. So is the rule that no AI output reaches a client or a platform without a person reading it first.
Plenty of agencies use AI and would rather you did not think about it. We would rather be direct: AI is integral to how Sublime Plus operates, and it is a large part of why we can move at the pace and price we do.
It is also, by policy, never the last step.
The rule underneath everything
AI is a tool, not a replacement for human judgement. Every AI output is reviewed, verified and approved by a human before it moves forward. Not sampled — every output.
That principle sets the shape of everything below.
Where AI does the work
Drafting. AI drafts posts, comments, replies, blogs and content plans. Every draft is reviewed by the Customer Manager before it is submitted anywhere.
Research and analysis. AI assists with client research, competitive analysis, SEO analysis and LLM audits. Anything it produces gets verified against primary sources where those exist.
Internal operations. Meeting prep, report generation, documentation, process work.
Editing and QA. AI helps with grammar, tone analysis and content review. It supplements the human review pass; it does not replace it.
We use Claude as the primary tool for content creation, research, planning and analysis. Adding a new tool to that list requires approval, and client work does not run through unapproved tools.
Where AI is not allowed near the work
The prohibitions are short and absolute:
- Never to generate fake testimonials, reviews or fabricated data.
- Never to produce content that breaches our ethical guidelines.
- Never to publish an AI-generated statistic without independent verification.
- Never to process client-sensitive information through an unapproved third-party tool.
The third one is worth dwelling on. A model will produce a confident, specific, wrong number as readily as a right one, and a fabricated statistic in a client's content becomes the client's problem the moment someone checks it.
Accountability sits with a person
Whoever submits, approves or posts a piece of content bears full responsibility for it, whether AI was involved or not. "The model wrote it" is not a defence available to anyone here, which is precisely the point of stating it that way.
Where we disclose it
On our AI influencer and blogger accounts, every bio and every post description states that the content is AI-generated, and platform AI labels are applied wherever they exist. That is not a courtesy — sponsored content from an AI persona that presents itself as a real person is exactly the kind of misrepresentation our ethical standards prohibit.
More broadly: we do not create fake grassroots movements, we do not misrepresent where content came from, and we do not fabricate information, statistics, testimonials or claims. AI makes it trivially easy to do all three quickly. That is the reason the rules are written down rather than assumed.
Want this run on your brand?
We will walk you through what the models currently say about you, and what it would take to change it.