12/10/2025
ARTIST STATEMENT: AI, CREATIVE PRACTICE, RESPONSIBLE STEWARDSHIP, AND PERSONAL PRACTICE
Artificial intelligence is not a neutral tool in the old sense. It is a culture scale force that remakes how images, sounds, texts, performances and ideas are produced, distributed, valued and policed. For working artists of every discipline this is a moment of both opportunity and risk. It is also a moral test. In plain language: AI is helping and it is hurting. How it ultimately changes creative life depends on the rules we insist on, the markets we refuse to normalize, and the habits we teach ourselves and others.
I speak as an artist who has both engaged deeply with AI and resisted its worst impulses. I have integrated AI into my practice thoughtfully, deliberately and transparently. The foundation of my output originates in my hand, my eye, my voice, my archive and my lived experience. AI has been, in my practice, an editor, a production assistant and a logistical partner. It is not the source of my creative identity. When AI offers textual edits or phrasing, those are suggestions. I always remain the progenitor of the words, the ideas and the decisions. In visual work I train Photoshop engines and other generative tools on my own imagery, my own garments, my own textures and my own photographs. That is critical. The results are extensions of my archive and my authorship, not anonymous imitators of other creators.
How AI helps artists
1. Faster iteration and expanded experimentation. AI lowers the friction to try new directions. Visual mockups, melodic sketches, choreographic cues and rough drafts can be produced quickly. That accelerates ideation and lets artists test risky choices affordably. 2) New forms of collaboration. Generative models can suggest variations an artist may not have imagined, surfacing hybrid aesthetics and new vocabularies. For many artists this expands expressive range. 3) Accessibility and scaling craft skills. Tools help artists with disabilities, low budgets or limited training to realize visions that previously required expensive teams or years of practice. 4) Productivity and business support. Automation of mundane tasks such as transcription, metadata tagging, formatting and outreach frees time for high value creative work. For small studios and independent artists this can be liberating. 5) New economic and exhibition possibilities. AI enables generative artworks, interactive installations and personalized creative experiences that can open markets and new forms of audience engagement. 6) Clarity and precision in language. Text assistants help with business planning, grant copy, professional correspondence and marketing materials. They do not create purpose. They help articulate and sharpen what already exists in the artist's mind.
How AI is hurting artists
1. Uncompensated copying and training on artists labor. Many large models are trained on works scraped from the web without consent or compensation. This reproduces styles and details from living artists and sells derivative outputs as if they were new. That displaces labor and undermines income streams. 2) Devaluation of craft and expertise. When commoditized AI outputs flood marketplaces they depress prices and condition buyers to expect acceptable results for the lowest cost. Years of craft can be priced out of existence. 3) Attribution and provenance loss. Generative outputs rarely carry reliable provenance. Audiences cannot distinguish human authorship from AI synthesis and that erodes reputational capital and trust. 4) Power asymmetry and platform capture. A handful of corporations control the most powerful models and distribution channels. Artists have little leverage over how their work is used, how revenue is shared or how decision systems prioritize content. 5) Bias, stereotyping and cultural flattening. Models inherit biases from training data and reproduce harmful representations. Minority aesthetics are misused or flattened into tokenistic caricatures. Context and cultural labor get stripped away. 6) Legal and psychological harms. Copyright law lags behind technology and legal remedies are expensive and slow. Artists face emotional harm when their signature work is parroted by a machine or when the marketplace privileges cheaply produced imitations. 7) Environmental costs. Training and serving large models consumes nontrivial energy and has ecological consequences. 8) Erosion of craft training. When inexperienced creators rely on AI before learning fundamentals, long term craft erodes. AI can simulate taste but it cannot substitute for lived, tactile mastery.
Intent matters but intent is not enough A user or corporate intent can be positive or negative, but stated intent does not eliminate harm. An artist may use AI purely to generate variants for a concept. A corporation can use the same workflow to pump out derivative products at scale and license them for profit. Both outcomes can harm living artists. We must judge actions by outcomes as well as stated motives. That means measuring who benefits, who is paid, who is erased and who bears the costs.
Principles for an artist centered approach
1. Consent. Artists must be able to opt out of having their work used to train commercial models. Opt out mechanisms must be real, accessible and enforceable. 2) Transparency. Models and platforms must disclose training sources, the percentage of copyrighted material used and whether outputs were partially or wholly AI generated. 3) Attribution. Outputs derived from identifiable living artists work should disclose that lineage and include a credit line. 4) Compensation. When a model uses an artist s work in training there should be mechanisms for fair compensation, whether collective licensing, revenue shares or statutory remuneration. 5) Control and provenance. Systems should embed provenance metadata that is tamper resistant so audiences and buyers can see authorship history. 6) Redress. Artists must have low cost, quick pathways to challenge misuse, demand takedowns and seek remediation. 7) Ecological responsibility. Model deployment should account for energy use and adopt offsets and efficiency measures. 8) Cultural stewardship. Communities must retain agency over cultural artifacts and indigenous or marginalized stylistic traditions must not be exploited without consent.
Practical actions artists can take now
1. Document and register. Keep records of work, dates, galleries, sales and social posts. Copyright registration strengthens legal standing. 2) Watermark and provenance. Embed visible or metadata based provenance where possible and use time stamped uploads. 3) Control distribution. Limit public exposure of high resolution masters that could be scraped and password protect archives when possible. 4) Educate clients and audiences. Explain differences between AI assisted and human made work in contracts and public materials. 5) Negotiate contracts. Insist commissions define AI use and demand attribution and compensation where appropriate. 6) Join or form collectives. Collective action increases bargaining power for licensing, legal defense and shared infrastructure. 7) Diversify income. Lean into experiences, teaching, limited editions and commissions that are harder to automate. 8) Learn the tools on your terms. Master AI as another medium when it suits your practice but do not let it substitute for the practices that define your voice. 9) Public advocacy. Support legislation and platform policies that require transparency, consent based dataset creation and equitable payment models.
Policy and platform asks
1. Mandatory dataset audits. Platforms should publish audited inventories of datasets used to train models. 2) Compulsory opt out registry. A publicly searchable registry where artists can exclude works from training corpora. 3) Fair remuneration frameworks. Collective licensing similar to music licensing for recordings with payments administered by artist organizations. 4) Provenance standards. Enforceable metadata standards that travel with files through distribution and sales. 5) User facing disclaimers. Algorithms that influence discovery and sales should flag AI generated or AI assisted works clearly. 6) Rapid dispute resolution. A low cost, independent arbitration process for alleged misuse that can order takedowns and compensation quickly. 7) Research funding for alternatives. Public support for open, artist governed models that prioritize ethical use and broad access without corporate capture.
How funders, galleries and institutions should respond
1. Adopt ethical acquisition policies. Do not exhibit or sell works derived from living artists without consent or payment. 2) Fund artist research. Provide grants for artists experimenting with AI and for artist led infrastructure for provenance and rights management. 3) Curate with context. When showing AI work include labels that explain authorship and methodology. 4) Support legal clinics. Partner with legal clinics that help artists navigate disputes and rights claims.
A model of responsible AI in my practice I integrate AI into my work in three concrete ways that model responsible practice. First, I use original source material. I train models on my own photographs, my garment archive, fabric scans, sketches and studio photography. This keeps outputs tethered to my authorship and prevents the use of unlicensed third party images. Second, I use AI as editor not origin. The creative spark remains human. AI helps with iteration, refinement, composition and finishing touches but it does not invent the underlying concept. Third, I practice transparency. When AI contributes to a piece or when it has helped in the editing or marketing process I disclose that fact. Transparency protects provenance and clarifies authorship.
How I use AI specifically Textual tools: I use text based assistants for business planning, grant applications, investor communications, scheduling and editing artist statements or descriptions. The tools suggest phrasing, tighten language and help structure arguments. I always choose which edits remain and which are discarded. Visual tools: I use Photoshop model training with my own imagery and words. I curate a training set composed of photographs I made, garments I sewed, fabric swatches I scanned and textures I created. By training on my archive the resulting generative outputs remain variations on my lineage rather than appropriations of others. Advertising and promotional images: When I create campaign images for advertising I build compositions from assets I own. AI assists in proposing compositions, color palettes, layout variations and text overlays. I then select, edit and refine the chosen outputs. The final ad images are products of my direction, my imagery and my approval. Business planning and editing assistance: AI helps me draft planning documents, generate marketing calendars, refine pricing language and format investor decks. It speeds non creative tasks and lets me focus on craft and big picture strategy.
Concrete examples that show the difference between exploitative and responsible use Exploitative use: Training large models on scraped high resolution images, then selling derivative prints that mimic living artists without consent, without provenance and without payment. Responsible use: Training models only on an artist s own archive or on properly licensed materials, disclosing AI involvement, and using the tool to scale reach while protecting the origin and integrity of the work.
The long form critique and the personal reflection together The deep critique includes legal, economic and cultural prescriptions. Law and regulation must catch up and provide opt out rights and remuneration where appropriate. Technical infrastructure must be built to carry provenance metadata. Market innovation must create revenue sharing frameworks so artists whose work materially shaped model outputs receive compensation. Cultural shifts must occur so audiences and institutions value craft, provenance and human origin as part of artistic worth.
At the same time the personal reflection shows that AI can be harnessed to preserve creative space. By offloading administrative tasks, generating organized options and providing editorial suggestions AI allows an artist to return to embodied work, to the slow refinement of material technique and to risk taking that machines cannot do. I am both clear eyed about harm and pragmatic about the real benefits I derive when I use the tools under the rules I set for myself.
Policy and structural recommendations summarized
1. Structural reform. Adapt copyright frameworks to address training on copyrighted material, create statutory remuneration mechanisms and require provenance. 2) Technological fixtures. Build opt out registries, portable provenance metadata standards and open models governed by non profit consortia. 3) Market innovation. Create licensing markets and revenue share systems so that artists are paid when models benefit from their work. 4) Cultural shifts. Revalue time, craftsmanship and intentionality. Educate audiences and curators to prize provenance and authorship.
Final practical checklist for advocates, institutions and artists
1. Document everything. Keep date stamped records and registrations. 2) Control master files. Limit exposure of high resolution masters online. 3) Ask for provenance and credit in all contracts. 4) Negotiate AI clauses in commissions. 5) Join collectives for bargaining power and shared infrastructure. 6) Support policy reforms and artist led models. 7) When in doubt, choose transparency and assert authorship.
Conclusion If one thing is remembered, let it be this. AI will change what is possible. It must not be allowed to change what is fair. The worst outcome is normalization of a market where AI derivatives substitute for human made work without consent, credit or payment. That outcome hollowed out cultural labor and concentrated creative capital in corporate hands. The best outcome is pluralistic. AI enriches human imagination while protecting the labor that makes culture meaningful. Achieving that future requires action on legal, technical, market and cultural fronts.
For artists: document your work, control distribution, negotiate contracts and organize collectively. For platforms and companies: be transparent, pay when you profit from other people s creativity and build means for provenance and redress. For policy makers: enact opt out rights, fair remuneration mechanisms and enforceable provenance standards. For funders and institutions: support artist led technology, legal defense, education and models that redistribute value rather than concentrate it.
This is not a call for technophobia. It is a call for stewardship. Technology without stewardship will automate away the human labor and cultural context that give art its meaning. We can choose a different path. We must choose it deliberately through policy, practice and solidarity. Artists should treat AI as a medium to be mastered and as a field of struggle to be shaped. That doubled commitment is the only honest way forward.
I suggest changes in my writing when AI helps me edit or clarify text, but I am always the progenitor of the words, the ideas and the creative will. AI proposes options. I choose, refine and claim authorship.
Am I taking anyone's job? No, I have always done all of the graphic and written output for my businesses. I was never going to hire anyone to do that work. Am I stealing from other people's output? No, I have learned to use AI by feeding it my own work...I can not steal from myself.
