No AI training
Any use of this website — including its text, code, images, designs, audio or visual content — to train, fine‑tune, or otherwise feed generative artificial intelligence (AI) technologies is expressly prohibited. This includes attempts to imitate our style or to generate text, code, images, audio, video or any emerging media. No exceptions apply without explicit written permission from Myriam Studios. For technical enforcement, see ai.txt.
What we mean by AI
Our focus here is on generative AI technologies such as large language models (LLMs), diffusion models, and similar tools that attempt to replace creative tasks. We stand firmly against exploitative practices such as training on scraped art or imitating the style of living creators, while we recognize and value the role of machine learning when it is applied responsibly to advance human well‑being in fields such as medicine, healthcare, and scientific research. Such applications demand the highest standards of ethical responsibility, but they are distinct from generative tools that threaten the integrity of creative practice.
Authenticity of our works
Myriam Studios never engages in deep fakes, process fakes, style theft or mimicry. Our portraiture is created by hand, reflecting the skill of achieving a likeness through direct observation. When we make homages, they are the result of painstaking research, our own analysis, and our own brushstrokes. We do not use AI in any part of our drawing or painting processes. We do not use AI to generate finished images, music, storytelling or anything we claim to be creative or authored. We do not believe art is just in the idea or act of curation. We see end to end prompt pipelines as anathema to artistic practice and a sure way to increase the amount of slop in our information space. We don’t do it. Every work we present is an artifact of authentic human effort and respect for the integrity of artistic practice.
Emerging ai tools
We recognize that AI technologies for creative workflows are developing rapidly and being normalized as part of industry‑standard software. Their scope is often unusually broad and as of 2025, the target task unusually fuzzy, compared to traditional hard‑coded plugins and software features. But it is impossible not to recognize that with enough control‑providing improvement, these technologies when trained on ethically licensed art archives can become invaluable in creating large‑scale projects that hereto would have been prohibitively expensive and time‑consuming for an indie studio.
Just as they can become slop‑generation slot machines exacerbating environmental pollution in our actual lands alongside our sensory and mental scapes—lest we as a culture develop a new form of media literacy and rediscover and weight the value of human thoughtfulness over algorithmic engagement.
Over the next few years, we see as the paramount task for our studio the project of developing a creative practice that stands immune to the attraction basins of easy slop and psychosis‑inducing recursive loops of meaning phantoms that come part and parcel with the sheer throughput of “AI‑assisted creativity.” Having lived in not just pre‑AI but pre‑internet era—a cultural vantage point that is bound to become extinct, we have a commitment to seeing the sentiment and depth, the warmth, the mental strain and intuitive virtuosity of creative activity and appreciation carry through the machinization of knowledge work and creative workflow retooling.
In light of this commitment, Myriam Studios monitors developments in AI closely and considers particular use cases with care. Where we choose to integrate and study AI, we do so judiciously and with full transparency.
Transparency
We make selective use of AI tools as a means of reducing tedious work and to assist us in research (the way a writer would use that word, rather than a scientist). We employ LLM-based AI text tools in areas such as web development and administrative copy generation, where they help streamline routine tasks: this policy was written with the help of Copilot.
Where visual AI tools are considered, they are limited to non‑creative technical domains — for example, retopology, rendering optimizations, or other forms of repetitive digital labor that do not constitute creative authorship or imply physical presence in a moment in time as part of the value of the work. In exploratory contexts, AI may occasionally be used to test parameters and assist in research at ideation stage. Even there, we don’t outsource ideation itself to an LLM. Instead we treat AI assistance in ideation as a method of interrogating world’s knowledge and idea encoding — so mostly as a conversational research tool — prohibiting it from reversing roles and spitting out ready-made “idea” prompts for us to execute.
In short, we distinguish between works meant to touch humans directly — such as drawing, painting, music, poetry and storytelling, authorship — which remain entirely human, and tasks that communicate primarily to devices or systems — optimization steps, formatting, or code generation — where AI may be used judiciously to reduce non‑creative burdens. We also recognize that certain communication tasks like social media marketing are primarily algorithmic once a brand and voice are established. In these contexts and in order to not burn out creating bespoke social media posts that have a shelf-life of five minutes, we might turn to AI for assistance, but never to create a false impression of human connection.
This policy will be updated as needed to reflect new practices and to ensure that our approach remains consistent with our values and our commitment to protecting human creativity.