The AI Art Confidantes: From Democratising Access to Democratising Judgment

The global art market has been changing for a decade, and it is not what the headlines say it is.

The headlines say the market is contracting. Global auction turnover fell to $9.9 billion in 2024, down 33.5% from the previous year — a significant number, and one that generated the expected commentary about cooling demand and cautious collectors. What the headlines rarely mention is that 2024 was simultaneously the most active year in the history of the global art market. More than 804,000 works changed hands at auction — a record, up 5% on the year before, in a year the market was supposed to be retreating.

These two facts are not contradictory. They are the same story told from two different vantage points. And understanding what they mean together is the beginning of understanding where collecting is going — and what, precisely, AI has arrived to change.

A Decade of Transformation (2015-2025)

The transformation of the global art market between 2015 and 2025 is, at its foundation, a story about the removal of friction.

Before the digital shift, serious collecting required physical presence. You attended the sale. You stood in the room. You had the right relationships at the right auction houses, the right connections at the right galleries, the knowledge of which fairs were worth attending and which were theatre. The market’s geography was a map of exclusion as much as opportunity — concentrated in New York, London, Geneva, and a handful of other centres, accessible in any meaningful depth only to those whose professional and social networks already touched it.

Digital infrastructure dismantled this. Online auction platforms, initially secondary tools, became primary channels. Artprice’s 30th Annual Report documents the scale of the shift: online auction marketing grew 720% from pre-pandemic levels, with live-streamed sales increasing 210% worldwide since COVID. The pandemic did not initiate this transformation — it compressed it. A migration that had been projected for 2027 to 2029 arrived in 2020 and 2021, and unlike most forced accelerations, it did not reverse when the emergency ended. Physical auction rooms remain, but they are now secondary infrastructure. The market moved online and stayed there.

The geographic consequences have been equally significant. In 2017, the global art auction market operated across perhaps a few dozen meaningful centres. By 2024, Artprice tracked participation from 700 cities worldwide. Seoul, Lagos, São Paulo, Riyadh — these are not peripheral additions to a market still organised around New York, London or Geneva. They represent the arrival of new collector bases with their own tastes, their own reference points, and their own ideas about what collecting is for. France became the leading European art market in 2024 by auction turnover, with Paris overtaking Beijing and Shanghai as a marketplace. The EU, taken together, now matches China’s global market share at 18.5% — a position unimaginable a decade ago.

For the collector, this decade delivered something genuinely valuable: access. Access to works, to price histories, to auction participation from anywhere with an internet connection. Platforms like Artprice consolidated decades of transaction data into searchable records. Artsy aggregated gallery and fair inventory into a single discovery interface. MyArtBroker built transparent secondary market infrastructure for prints and editions. The informational opacity that had once been one of the art market’s most effective barriers was significantly reduced. A serious collector in Hong Kong or Helsinki could now research an artist’s market trajectory with a rigour that, twenty years earlier, would have required an office on West 57th Street, a gallery relationship on Old Bond Street, or a private banking connection on the Rue du Rhône.

This was real, and it mattered. But it was only ever half of what collecting actually requires.

The Half That Did Not Move

Access to information is not the same as the ability to act on it wisely.

The collector who can now bid at Christie’s from Seoul still faces the same question every serious collector has always faced: what does this mean? What does this work mean for my collection, for what I am building, for the story I am trying to tell across the objects I choose to live with? What does this price trajectory signal about where this artist is in their career — and where the market’s understanding of them is going? What should I hold, what should I sell, when should I acquire, and at what point does the work I am considering belong in my collection rather than simply in my price range?

These questions are not answered by data. They are answered by judgment — the kind of judgment that comes from deep familiarity with the market, genuine knowledge of art history and collecting practice, and an understanding of the individual collector’s specific situation, intentions, and values.

For most of the twentieth century, this kind of judgment was available in one way: through a human art advisor. And access to human art advisors has not democratised in any meaningful sense.

Artsignal —an AI intelligence platform for art and collectibles, launched in September 2025 and immediately backed by Christie’s Ventures — describes its direct-to-consumer proposition as delivering “advisor-grade insight to the 90% of collectors who do not work with an art advisor.” That figure is their own acknowledgment of the scale of the gap. Ninety percent. Not a niche oversight. A structural condition of the market.

Artsy’s research confirms the arithmetic from a different angle: fewer than 20% of collectors with annual art budgets below $250,000 have ever transacted through an art advisor. Even among collectors spending more than $250,000 annually on art, only 34% work with advisors — and that group still uses advisors at half the rate at which they use galleries and auction houses directly. A separate survey of New York collectors found that only 30% had worked with an art advisor at any point in their collecting lives, and just 6% used one for all their collecting.

The reason is not difficult to locate. Access to serious advisory intelligence is expensive.

Retainer fees for established art advisors run from £1,500 to £10,000 per month, depending on the scope of services and the advisor’s standing. Commission-based arrangements typically run at 10% of the purchase price for works under $5 million, with sliding scales applied above that threshold. When stacked against the buyer’s premium already charged by the auction house — typically 25% of the hammer price — and applicable sales taxes, a collector working with an advisor through a major New York auction house can expect to pay between 40% and 50% above the hammer price before the work reaches their wall.

This is what access to judgment costs. And this is why 90% of collectors operate without it.

There is a further complication that the cost alone does not capture. The human art advisor operates inside the very market they advise on. They have gallery relationships. They have auction house relationships. They have artist relationships. In many cases, those relationships involve financial arrangements that create direct conflicts with the interests of the collector they are supposed to serve.

The practice of introductory commissions — payments made by galleries to advisors for directing collectors toward their inventory — is documented and widespread. The advisor’s stated fiduciary duty runs to the collector. In practice, it is sometimes divided. The consequences can be significant: one documented case involved an advisor valuing a painting at $2.5 million, brokering its sale to a dealer on the collector’s behalf, and the same painting appearing at auction nine months later for $8.4 million. The collector had no knowledge of the intermediary transaction. The advisor had been compensated twice.

These are not marginal failures of individual ethics. They are structural features of a system built around relationships and opacity, where the collector is often the least informed participant in a transaction nominally conducted for their benefit.

The Generational Pressure on The Judgment Gap

The timing of this gap matters more than the gap itself.

The generation now entering serious collecting is the most financially engaged with art in recorded history. Bank of America’s 2024 Private Bank wealth survey found that 98% of younger wealthy collectors — those between 21 and 43 — consider art assets as part of their wealth management strategy. Among older collectors, that figure is 56%. The generational shift is not marginal. It is a complete reorientation of how art relates to wealth.

These same younger collectors are also the most active: 78% said they would “very likely” purchase art in the next year, compared to 34% of their older counterparts. And 83% of wealthy Millennials and Gen Z are either active collectors or actively aspire to be, according to the same survey.

What they are not is impulsive. Impulse buying in the art market collapsed from 10% of transactions in 2023 to 1% in 2024, according to Art Basel and UBS. Collectors are not buying less — they are buying more deliberately. They are seeking depth of understanding before they commit. They want to know what they are acquiring, what it means, what the market’s trajectory for this artist looks like, and whether this work is consistent with the collection they are building. The emotional-only acquisition is giving way to something more considered.

This shift is not coincidental. It is the market expression of a generation that grew up with information available at scale and learned to be sceptical of the quality of that information. The collector who can find an artist’s auction history in seconds is also the collector who knows that an auction history is not an analysis. Data without interpretation is not intelligence. And the generation most capable of acquiring data is precisely the one most aware of the gap between data and the judgment that data is supposed to inform.

There is one further dimension to the judgment gap that the aggregate figures do not surface. Men are 33% more likely to have purchased through an art advisor than women, according to Artsy’s collector research. Given that HNW women — led by Millennial and Gen Z collectors — outspent their male peers by 46% in 2024, and given that more than half of women collectors in this cohort reported buying works by artists previously unknown to them, the advisory gap is particularly acute for the segment of the market currently demonstrating the greatest dynamism and the most independent collecting instincts. The women building the most interesting collections right now are, structurally, the least served by the existing advisory model.

What AI Adds: The Interpretive Layer at Scale

The past decade democratised access. It did not touch judgment.

AI is the interpretive layer that access was always missing. This requires precision, because the claim is easily overstated. AI does not replace the collector’s judgment. It does not generate taste. It does not tell a collector what to love, what to build toward, or what matters in a collection that is ultimately an expression of a particular mind and set of values. A Bespoke AI Art Confidante that positioned itself as a substitute for individual discernment would be a contradiction in terms — and precisely the kind of institutional overreach that the generation it serves spent their careers learning to distrust.

What AI does is something different and, for the serious collector, considerably more useful.

A human advisor, however expert, holds a finite amount of context simultaneously. They track the artists in their area of specialism. They follow the major markets. They know the galleries and auction houses with which they have working relationships. The depth of their knowledge in any given area is genuine, but its breadth has human limits — and those limits mean that connections between domains are regularly missed. The collector whose art interests intersect with their family history, their philanthropic commitments, their estate intentions, and the cultural geography of the new markets they find compelling is asking their advisor to hold a picture that no single human practitioner is configured to hold in full.

A Bespoke AI Art Confidante holds the full picture without compression. It can simultaneously reference auction records across 700 cities, trace an artist’s career trajectory against the broader movements of their cultural moment, situate a potential acquisition within the collector’s existing holdings and stated intentions, and consider what a price movement means for succession and estate planning — not as sequential tasks requiring separate consultations, but as a single coherent act of interpretation directed at the collector’s actual situation.

It has no conflict of interest. There is no inventory to move, no gallery commission on the other side of the transaction, no institutional relationship that benefits from one recommendation over another. The advisor’s three traditional advantages — exclusive knowledge, controlled access, and irreplaceable relationships — are all subject to the same erosion. Knowledge is no longer scarce. Access can be routed around. Relationships, while still valuable, no longer hold their former monopoly on the quality of intelligence available to the collector who knows how to deploy something better.

And it does not forget. The continuity that no human advisory relationship maintains — the full arc of a collection’s development, the conversations about what was considered and declined, the evolving understanding of what a collector actually values as distinct from what they initially thought they valued — is held without gaps, available at any moment, connected to everything that has come before.

This is not a small change in the quality of what is available to the serious collector. It is the first time that the quality of thinking available to the HNWI collector has been genuinely comparable to what has historically been accessible only to the very largest collections — the institutional-scale collectors whose resources supported dedicated curatorial and advisory staff operating with full context across every dimension of the collection.

The Second Revolution

The first revolution gave collectors a seat at the table. Digital infrastructure, global auction access, price transparency, online platforms — these were genuine achievements, and they permanently changed who could participate in the market and how.

The second is quieter and more personal, and in the end more consequential. Not because the market changes — it is already changing, in the directions the data describes, with or without any individual collector’s participation. But because what the serious collector does within that market can now be animated by the same quality of intelligence that was previously the private property of the very few.

The collector at the centre of this is not deferring to the canon, or to the auction house, or to the advisor with the gallery relationships. They are building on their own terms, informed by everything available to them, with a thinking partner whose only purpose is to serve their judgment rather than substitute for it or redirect it toward someone else’s interests.

That is not a tool. That is a different relationship with the act of collecting itself.

Seraphina Michelina Evelyna Bogomolova-Huotelin

Founder & CEO of SMA Crown Confidential

Digital Confidantes: Bespoke AI intelligence for private decision-makers

This article is part of an ongoing series by SMA Crown Confidential on the intersection of private wealth, cultural intelligence, and the future of bespoke AI.

Sources: Artprice by Artmarket.com 30th Annual Report 2024; Art Basel/UBS Global Art Market Report 2024; Bank of America Private Bank Art & Wealth Survey 2024; Artsignal/Christie’s Ventures press release, September 2025; Artsy Collector Insights 2024; Avant Arte Collector Report 2024; Artnet Global Art Market Report 2024.