At some point last year I stopped arguing with the premise and started asking a different question. The premise I kept hearing, at trade events and in podcast interviews and occasionally from other small producers, was this: AI is fine for big industry, but craft is different. Craft is human. The two do not mix. I have heard versions of this argument more times than I can count, and I have come to think the tension is slightly false. Not because AI is some neutral force that poses no risks to small producers, but because the framing misunderstands what craft producers are already doing every single day.
We use calculators to work out ABV. We use Excel spreadsheets to formalise recipes so that the rest of the team can reproduce them. We back our records up to the cloud. We photograph our botanicals for Instagram, edit the images with software, and schedule the posts. We track sales in a database and send purchase orders via email. None of that is controversial. Nobody stands at a trade show and says, "You used a calculator to check your ABV -- that's not craft anymore." The calculator is invisible. It is accepted. It is just a tool that accelerates a process we would otherwise do by hand, and doing it by hand would not make the product better. It would just take longer.
AI is the same category of tool. That is the argument, and I think it holds.
The digital infrastructure craft businesses already run on
Every working craft producer in 2026 is running on digital infrastructure. This is not a criticism; it is a description of how small specialist businesses actually function. Asterley Bros operates out of a South London workshop with a team of six. Our recipes are versioned documents. Our production logs are digital. Our compliance records sit in cloud storage. When I need to scale a batch or convert a unit, I use software. When I need to communicate a recipe change to the team, I write it up and share it via a collaborative document.
Circumstance Distillery in Bristol, run by the talented Ginette de la Fare, is one of the most considered craft operations in the UK. They are meticulous about provenance, about local botanicals, about the narrative behind each release. They are also, inevitably, running production software, using digital records, and managing their brand through digital channels. The craft identity and the digital tooling are not in conflict at Circumstance. They coexist, because they have to. The business would not function otherwise.
According to Craft Spirits Magazine, the majority of American craft distilleries now use some form of distillery management software for production tracking and compliance. That number has grown steadily as the compliance burden on small producers has increased. The software does not make the spirit. It makes it possible to produce the spirit while also running the rest of the business.
What AI actually does in a small drinks business
The version of AI that matters for most small craft producers right now is not the version that generates spirits from a prompt or replaces the palate of a master blender. It is considerably more mundane, and considerably more useful. It is the layer that accelerates data collation, surfaces patterns in production logs, drafts internal documentation, and disseminates decisions to the team in a format they can use.
At Asterley, we have been building internal tools with Absolution Labs that sit in exactly this layer. A tool that takes a tasting note and formats it into a spec sheet. A process that pulls our sales data and flags which SKUs are moving and which are sitting. A draft compliance document that we review and approve rather than compose from scratch. None of these tools make a single craft decision. They reduce the time we spend on tasks that do not require craft expertise, so that more of our hours go toward the ones that do.
I want to be careful here about scale. Bespoken Spirits in California is using AI at the other end of the spectrum, applying machine learning to accelerated spirit maturation and flavour modelling across large production volumes. Their work is interesting and technically significant. But it is also operating in a fundamentally different context from a team of six making small-batch aperitivo in South London. The question for craft producers is not "should we do what Bespoken Spirits does?" It is "which parts of our operation consume time without contributing to the product, and can AI address those parts?" Those are very different questions with very different answers.
Where we draw the line: what stays human at Asterley
The comparison below reflects how we actually think about this inside Asterley Bros. It is not a rigid policy. It is a working distinction that evolves as the tools evolve.
| Stays human | AI-assisted |
|---|---|
| Tasting and approving each batch | Logging and formatting tasting notes |
| Deciding which botanicals to source and from whom | Tracking supplier lead times and flagging shortfalls |
| Setting the flavour profile for a new expression | Surfacing comparable batch data from production records |
| Brand voice and creative direction | Drafting first versions of compliance and spec documents |
| Final call on what gets released | Generating sales and distribution summaries for review |
The pattern is consistent across every row. The decision that matters -- the one that shapes what the product is and what the brand stands for -- is made by a person. The AI handles the surrounding work: the formatting, the surfacing, the communication, the administrative layer. This is not a compromise of craft. It is craft operating more efficiently.
Tastry in California has been working on AI-assisted sensory analysis and consumer preference prediction for some years now, and even in that more sophisticated application the output is a recommendation that a human evaluates against their own palate and brand judgement. Tastry's tool does not decide. It informs. That distinction matters.
The argument worth having
I am not interested in the argument that AI cannot be used in craft businesses. That argument has already been lost by the people making it, because every craft business already uses digital tools that would have seemed equally alien twenty years ago. The interesting argument is about where to draw the line: which decisions should remain entirely with the maker, which can benefit from AI assistance, and which are purely administrative and can be fully automated without any loss to the product or the identity behind it.
That is a conversation worth having between the maker and the technologist, ideally by someone who sits on both sides of this AI and drinks manufacturing divide. At Asterley, we are working through it batch by batch, tool by tool, in the margins of a production schedule that does not slow down to accommodate philosophical uncertainty. The calculator was always there. We just did not spend much time arguing about it.
Frequently asked questions
Does using AI mean a craft product is no longer handmade?
No. AI handles data collation and communication tasks -- the same category of work a spreadsheet or calculator already does. The tasting, blending, and production decisions remain with the maker. Using AI to format a recipe for your team does not change who made the product.
What kinds of AI tools are actually useful for a small craft distillery or drinks producer?
Internal tools tend to deliver the most value first: automated production logging, AI-assisted recipe formatting, demand forecasting from sales data, and compliance document drafting. These are tasks that consume time without contributing to the product itself. A team of six can recover significant hours per week from this category of automation alone.
Can AI replace a master blender or head distiller?
Not in any meaningful craft context. Companies like Tastry in California use AI for flavour prediction and sensory analysis, but even there the output is a recommendation that a human evaluates. The craft decision -- what this batch should taste like, what the brand stands for, what gets released -- is not a data problem. It is a values problem, and that stays human.
How does Bespoken Spirits use AI differently from a craft producer?
Bespoken Spirits uses AI-driven accelerated ageing and flavour modelling at significant scale, optimising for speed to market and product consistency across large batches. A craft producer like Asterley Bros is working at the opposite end: small batch, high manual involvement, and a flavour profile that is deliberately not optimised -- it reflects specific decisions made by specific people. The tools can overlap; the intent is different.
Where does Absolution Labs fit into this?
Absolution Labs builds internal AI systems for drinks businesses -- the operational and administrative layer that sits behind production. The goal is to reduce the time a small team spends on tasks that do not require their craft expertise, so more of their working hours go toward the decisions that do. You can read more at absolutionlabs.com.