Dear Jacky:
As an illustration agency, what are you doing to support artists during this era of AI?

At Jacky Winter, we’ve always been interested in how technology can expand our artists’ reach and increase their revenue.

We built Cobbler, a bespoke software for building and tracking estimates and invoices; developed Replier, a formatting tool for unruly email chains; and launched an artist-first digital newsletter platform called PencilBooth.

An open-minded attitude to technology has always benefited us, but when it comes to AI, our curiosity is tempered by concerns about its ethical and environmental impact.

In recent years, we’ve educated ourselves about AI so we can, in turn, guide our artists through yet another era of rapid change. We’ve monitored copyright lawsuits, written articles advising artists on protecting their work against scraping, and joined panels and webinars to discuss the threat of AI within the creative industry.

Views vary dramatically from artist to artist and client to client. While some creatives understandably want no part in AI, others consider it a game-changing innovation, saving hours of tedious work and allowing the time for true creativity to flourish. With differing attitudes among our roster of artists, we aim to cultivate an open and non-judgmental space where our artists' rights are protected, and clients’ wishes are considered, while also allowing those interested in exploring AI’s potential to do so. But with so much in flux, how can we do this? 

In late 2024, we added the following clause to our terms and conditions:

The Client agrees that it will not use the Deliverables (or any part of them) for the purposes of training artificial intelligence (AI) systems unless the Usage Terms expressly permit such use.

By August of this year, it was time to expand on this. Both the technology and our clients and artists’ use of it had changed. A policy to outline our expectations of clients and guarantees from our artists around AI use was drafted, reviewed internally, shared with our artists for input, and eventually published on our website with a commitment to review and update it every three months.

Transparency is key. It felt essential that all final artwork, unless clearly stated or specifically briefed, would be human-made rather than generated. Outside of that, we chose not to limit how artists used AI tools within their creative process - scenarios such as sourcing references for different character poses, or extending artwork backgrounds needed to be considered, and AI is now baked into many creative programs - but our artists are required to explain if and how they’ve used AI, when requested.

We spoke to clients with company-wide rules banning the use of AI, so to avoid sticky situations, we ask that this be shared at the briefing stage, so that we can respect the client’s wishes from concept sketches right through to final file output.

As AI becomes commonplace in studios and agencies, clients might not realise how their use of AI can impact the artists they collaborate with.

Our policy is pretty clear: our artists’ work can’t be used to train AI models or modified with AI tools unless it’s discussed and we provide written permission. Recent conversations with clients showcase why this is a must.

We’ve seen situations where creatives fed artists’ work into an AI tool to generate illustrations for a brief. Sometimes this happened before the artist had even confirmed they were interested and available for the project. We can see the positives of this approach - it’s a quicker, clearer direction than a sketch, and it can sell a concept to an end client who might struggle to visualise illustration as part of their campaign. Art directors and designers may see the use of AI in the briefing process as win-win. If they plan to hire the artist for the final deliverables, does it really matter how the scamps are created?

But this approach has its downsides - for one, training an AI tool with an artist’s work means it’s trained forever. It could be used to generate any number of images in the future - all without the artist's consent. The same model will pull from other datasets to generate the briefing images. Data that has been scraped from photographers, designers and illustrators - many of whom also haven’t given their consent.

We also worry that these tight briefs don’t leave space for the exploration and concept development that lead to truly exciting creative work. If a client is generating images and simply asking the artist to create a polished version, they aren’t leaning into the artist’s unique viewpoint or problem-solving skills - and there’s a risk that the value of the artist’s involvement could be minimised.

In some cases, an artist's work was animated or re-formatted by the client using AI tools. This raised concerns about quality for the artists, and was especially disconcerting when the first time we’d heard about it was when the campaign went live.

The most surprising situation didn’t involve a client we were working with but an AI company that claims to be ethical and artist-led. They used an artist’s illustrations for a tech brand to generate images showcasing their AI’s ability to replicate that brand’s illustration style. It’s frustrating to see an artist’s work used without permission to promote a tool that is positioned to replace their job, especially when the company didn’t grasp why it was problematic that the artist hadn’t been consulted.

These situations underscore why we believe our AI policy should be front and centre, not buried in the fine print of our contracts. We aim to spark meaningful conversations with clients and the broader industry about partnering with artists in a respectful manner.

Despite the doom and gloom conversations around the damage that AI is doing to the creative industry, we see reasons to be cheerful.

Flooding advertising with AI-generated imagery could lead to a more visually aware general public. I see people studying advertisements on the tube for signs of AI use. If something doesn’t sit right with the visual content they consume, people are taking the time to work out exactly what feels wrong in a way that feels far more public than picking apart bad post-production in advertising in the days of yore.

I hope that a more visually aware public will push brands to see more value in creative work, leading to bigger budgets, better timelines and a stronger industry. Perhaps we’ll develop ethical ways that AI can be part of the creative process, not replacing artists but assisting them - and we would love to hear from brands who are curious about exploring these possibilities.

We can’t predict what will happen as the technology develops, but what we can do is educate ourselves and our clients, keep our heads out of the sand and be part of the conversation to help shape a creative industry that values both innovation and artistic integrity.