For many, storing fine art pieces is defined by stillness: A controlled room, a locked door, and the assumption that minimal movement meant minimal risk. Inertia by necessity.
This is changing. Not because art has become less fragile (in some cases quite the opposite!) but because expectations have changed. Chiefly, art now travels more frequently, to more museums and is viewed by more people, all of this is good for culture, but it means as specialists, we have to think hard about novel methods of handling it.
Ostensibly unrelated to all this, artificial intelligence has moved rapidly from theory into everyday conversation. Large language models (LLMs) have increased public awareness of AI’s potential and risks. Regulators, developers and cultural institutions alike are grappling with what this means in practice. As Dr Geoffrey Hinton, often described as the “godfather of AI”, has cautioned, it “demands responsibility as much as enthusiasm.”
But what does all this mean for fine art? Does AI actually have a role in facing these new expectations, or is it a case of trying to fit a square peg in a round hole? Are there areas where intelligent systems genuinely improve care?
Let’s explore.
Not automation for its own sake
The fine art industry has, surprisingly to many laymen we speak to, always been at the forefront of certain technologies. From advances in conservation science to improved transport engineering, there’s a host of logistical innovations that were first pioneered in our space.
This is already embedded through SmART/Tech in our business, which applies environmental control to protect and enhance collections. Using motion sensors and responsive climate systems, SmART/Tech automatically adjusts light, temperature and humidity based on real-time conditions in an artwork’s immediate surroundings.
This reflects a broader truth about AI in fine art handling. The most valuable uses are often quiet and unseen and crucially, they support conservation rather than replacing expertise.
Supporting the art ecosystem, not reshaping it
Beyond physical care, artificial intelligence is beginning to influence how art is researched, discovered, and understood. We usually associate this sort of work with our sister-company, Crown Information Management. However, Large language models can assist with cataloguing, background research, and education, helping curators and collectors navigate increasingly complex information landscapes. They also help analyse data that supports informed collecting or lending decisions, particularly when dealing with dispersed or hard-to-access sources (“where can we form the most mutually beneficial relationship to lend our items to?”)
What they cannot do, however, is replace the relationships and judgement that underpin fine art handling. Moving a work safely from A to B, assessing its condition, or advising on display and storage requires accountability, experience, and trust. We’re an industry founded on person-to-person level trust after all. AI can enhance these processes sure, but it cannot deliver them in totality.
Our approach emphasises this balance. New tools are first evaluated within our ecosystem for how well they complement existing services, not how dramatically they transform them. Sustainability analysis, for example, is an area where data-driven insights may help reduce environmental impact over time, but always within the framework of bespoke, human-led planning.
Caution, responsibility, and the limits of AI
The pace of AI development has rightly raised concerns around misuse. Several governments have already taken steps to restrict or review the use of certain AI platforms, particularly where personal or sensitive data is involved, though it remains to be seen what the “GDPR-like template” for AI will be moving forward. In the art world, there are additional risks around forgery, plagiarism, and market manipulation, all of which demand vigilance from all of us.
Equally important is recognising what AI cannot do. Large language models analyse and generate text, but they do not act proactively. They cannot physically move an artwork, inspect a frame, or respond to an unfolding situation on the ground. Nor do they yet offer clear value when deployed as bespoke systems for art handling, where complexity, cost, and limited returns often outweigh the benefits.
For these reasons, we remain deliberately selective. Technology is adopted where it clearly improves protection, efficiency, or experience, and set aside where it risks undermining the personal service clients expect.
A measured path forward
Artificial intelligence will provide new possibilities for the fine art sector. Virtual and augmented reality, for example, may reshape how audiences experience exhibitions in years to come, expanding access while preserving the original.
For now, the priority remains clear. Art deserves systems that are intelligent but accountable.