At Re, as at many studios, the way we work is evolving rapidly as tools like Midjourney, ChatGPT and Stable Diffusion become part of our creative process. Almost every day, practical questions are emerging about ways we could use AI, as well as ways we’d like to avoid. So, in discussion with our worldwide teams, we set out to create principles to guide our day-to-day work with clients. 

We’re proud to share version 1.0 of our studio’s approach to integrating AI into our practice. Our plan is to continually update these principles as society and technology advance, and our own understanding grows. If you’d like to comment or feedback, you can access the figma file here.

Real representation matters.

Images tell people’s stories. We won’t use AI as a way to bypass proper consultation and engagement with people, or to co-opt someone else’s narrative.

Recent examples, such as AI generated black model Shudu (whose creator is a twenty-eight-year-old white man) and Levi’s looking to ‘increase diversity’ by using AI images in online shopping show a world where genuine representation risks being lost. While we’ll use AI to mock up concepts, or fill in background detail, when it comes to representation, only real people will do.

It’s not cool to copy someone else’s work.

While we will draw inspiration from other people’s style, we never use a current artist, photographer, filmmaker or any creative’s name in a prompt unless it’s in obvious homage to their work and the reference would be clear to audiences.

We always disclose the use of AI where it’s central to the creative output.

We believe transparency and accountability will help build trust between users and AI systems. We will always be open about where AI is used in
our work, particularly when it’s been a fundamental part of the process.

Public concern and mistrust over the use of AI remains high. Amnesty International recently came under fire for using AI generated imagery in social media posts regarding human rights abuses in Colombia. Even though the images included text stating that they were produced by AI, and Amnesty International’s reasoning was strong – to protect protesters from possible state retribution, the public still felt misled. As we continue to define what’s acceptable and what isn’t, openness in the use of AI is key.

We believe AI operates in the service of human creativity.

AI is a great way to break free from a blank page and quickly visualise nascent ideas, but it’s only one part of the creative process. We are mindful of identifying the opportunities and efficiencies of AI, as well as its limitations. We are committed to continuing to work with a broad array of creative specialists, redefining our relationship with them to dedicate more time to the sort of thinking that only humans can do. Because only human ingenuity can push creativity forward and create work that is authentic and distinctive.

We never take AI at face value.

AI can only work with what already exists, and we don’t have to look far to understand the biases that are currently ingrained in AI data sets. If left to its own devices, for example, AI will depict CEOs and Creative Directors exclusively as white men. It can’t deviate from the mean, or push to make things better. Only we can do that. It’s our role to understand these technologies deeply and bring diverse human perspectives into the process as we strive to push beyond our own biases and shortcomings.

Only fools rush in.

We’re excited to embrace the opportunities and efficiencies of AI, but we’re also mindful of proceeding with care. Challenges such as AI’s potential biases, intellectual property issues, and data security and privacy have the potential to cause serious problems for brands. We commit to staying informed so we can help keep our clients protected.

It’s our aim to review and update these principles regularly. In the meantime, if you’d like to chat about how they apply to your business, get in touch: benjamin@re.design