Does the medium start to look like infrastructure?

In 2025, it seemed that ‘digital art’ became harder to describe in the old shorthand. Not because the work suddenly stopped being screen-based, code-based, or blockchain-linked, but rather because the most serious practices no longer behave like an isolated niche. They behave more like contemporary art now: interdisciplinary, material, argumentative, and increasingly entangled with institutions (ones that used to treat the field as a speculative side-room).

That shift, from novelty to something closer to cultural infrastructure, is a shared undertone across four conversations I’ve had with people who work on different edges of the market: Elio D’Anna, co-founder and creative director of House of Fine Art (HOFA) Gallery in London; Mimi Nguyen, founder of Nguyen Wahed; Adam Berninger, founder of Heft Gallery in New York; and Stephen Santoro, who runs Vertu Fine Art at the intersection of advisory, gallery practice, and what he calls “market literacy.”

Their vocabularies differ: one talks in terms of “systems” and “lineage”, another in “patronage” and “institutions”, another in the practicalities of “care”, “display”, and “education”, and another in “the economics of streaming” and “on-chain royalties”. But they all keep circling the same feeling: 2025 was a year of recalibration. The work didn’t get smaller; the frenzy around it did. In that contraction, certain assumptions collapsed, whereas other practices started to read as durable. Or, as Elio D’Anna puts it: “the biggest shift I’ve seen in 2025 is a move from hype to practice.”

From hype to practice

When Adam Berninger says he avoids the label ‘digital art gallery’ with Heft, which opened in April 2025, it’s a way to underpin a reality that has already arrived. “Digital technology is no longer a category but an ever-present condition of our world,” he says. “Artists today move fluidly between mediums, and many of the works we show aren’t ‘digital’ in any conventional display sense at all.” Berninger reframes the discussion away from display formats and toward the deeper questions: how artists think, how they build, what they critique. “What interests us is how artists engage with systems: computational, material, social, or conceptual – and how contemporary technologies are folded into those systems.”

That emphasis on the system as a whole, and not all single entities or a big, overarching art category, shows up across all the interviews. It also reshapes what the concept of “hybrid” (or “cross-over” or “interdiscipline”) means. D’Anna says he sees artists integrating “performance, sculpture, sound or installation” so the work becomes “genuinely interdisciplinary, not just ‘screen-based’.” Besides that, he describes generative and AI artists becoming “far more critical about their tools, datasets and ethics.” It seems like the work is no longer (or at least less) evaluated simply by how it’s made, but more by how it is trained, constrained, combined, situated, and thought of (for the future). That’s a hybrid form for you.

Mimi Nguyen, who straddles the commercial art world and academia (she teaches at Central Saint Martins), offers a useful time-lag model. She talks about reading cyberfeminism and cyborg theory “more than ten years ago,” then only now – “literally today!” – paying for the bus with an Apple Watch. That gap between theory, culture, and daily adoption matters because it explains why the art world feels both late and early at the same time. The ideas are old; the infrastructure is newly ordinary. “But I anticipate we will become even more native to the medium in 2026,” she says.

Saying ‘NFT’ has stopped and the institutions are listening

One of the most tangible shifts described across the conversations I’ve had is the rhetorical one: the slow disappearance of ‘NFT’ as the biggest lens. D’Anna notes “more institutional curiosity: museums and foundations engaging with digital art not just as ‘NFTs’ but as time-based, conceptual works that need proper conservation strategies.” This suggests a big pivot from transaction to stewardship, from what a collector “gets” and “can sell” to what an institution must maintain over decades. Nguyen grounds that shift in a set of exhibitions that have expanded public awareness of computer-based art: Harold Cohen at the Whitney, Vera Molnar at the Centre Pompidou, Electric Dreams at Tate, and Samia Halaby’s inclusion in the upcoming Whitney Biennial. She also points to how the fair circuit, in semi-commercial contexts, has helped extend these conversations, like Cohen also showing at Frieze Masters 2025.

However, Nguyen is careful not to oversell momentum. “The pace of change remains slower than many anticipated, largely due to current market constraints,” she says, describing how mega-galleries still prioritise “established, lower-risk works within large group booths,” which limits space for technically demanding practices. Her more optimistic signal comes from foundations “doubling down on complex, research-intensive digital work,” citing projects by Lu Yang or Jakob Steensen at the Louis Vuitton Foundation and Lawrence Lek and Lauren Lee McCarthy at LACMA. The implication seems structural to me: institutions can underwrite complexity when the market won’t – and if they keep doing it, the market eventually (and hopefully) learns how to follow.

Stephen Santoro, speaking from an advisory position, notices institutional confidence rising in more symbolic terms. “Institutional interest has risen significantly,” he says, adding that “you can see this in moments like the MoMA acquisitions, which signal a level of conviction in our world that feels meaningful. From our vantage point, it seems likely that this trend will continue to build.” To frame the past few years, he’s basically describing it as a compressed educational cycle: “We’ve speedrun art history and I think that's something to be very proud of.” Thus, it seems like the institutional record has thickened over 2025 and will thicken more throughout 2026. To the gallery owners that matters more than a season of sales.

The long game

If institutions supply legitimacy, fairs supply exposure. Nguyen describes a deliberately aggressive first year for her gallery with seven fairs in 2025: Art Brussels, Art Dubai, NADA Warsaw, Warsaw Gallery Weekend, Paris Photo, Art Basel Miami Beach, and NADA Miami. “For a gallery barely a year old, this was undeniably excessive,” she admits, but “it was a way to meet a lot of people, gather data, and better understand the mechanics of the fair circuit.” And actually, in 2026 she expects no major structural changes. “This is a long game, a marathon rather than a sprint.”

That language – long game, patience, refusing to “compromise curatorial quality in order to sell for the sake of selling” – runs against the caricature of digital art as velocity. It also lines up with Berninger’s founding impulse, which begins with a kind of negative clarity: “Almost everyone I spoke to advised me not to open a gallery.” He read that discouragement as “telling,” an indicator of “exhaustion with certain inherited models: how galleries operate, how access is controlled, and even the emotional register in much of what is being shown.”

For Santoro, one fair moment last year became a shorthand for a broader breakthrough: “Zero 10 at Art Basel Miami.” He calls it “an objective win all around, for our space, for our artists, and for the builders, writers, curators, and everyone who has been contributing to this ecosystem.” He sees Vertu as uniquely positioned: “Not as a newcomer chasing heat, but as an organization that has been actively curating within this world since 2021, and researching and engaging with it even before then.” In his telling, Zero 10 didn’t invent the field; it made the field visible to people who had managed not to look.

Nguyen also describes the “Zero 10 sector at Art Basel Miami Beach” as “a clear indication of a change, a shifting paradigm,” while stressing how hard it remains to shift the contemporary art world, with its “entrenched rules and hierarchies.” Still, she sees a new generation. “Not only collectors, but also researchers, advisers, curators, and dealers.” Nguyen also says that with technological innovation, adoption often follows a more exponential curve. “It is striking how much has changed even over the past year alone. If we avoid a major recession, 2026 could be a very interesting year for digital art.” That said, many factors remain outside the art world’s control. “Political, social, and financial conditions will inevitably shape how this field evolves. Digital art does not exist in isolation, and its future will continue to be entangled with much broader structural forces.”

I guess you could say that this is the sort of realism that 2026 (and further) will likely demand: the work can grow culturally even when the market is cautious, but it cannot pretend that macro conditions are irrelevant.

The (return of the) object

An interesting consensus that came forward across all four interviews is that the future of digital art does not look purely digital. Not because artists are retreating into physicality as a way of saying that an online collection doesn’t ‘work’, but because it’s important that the art is being encountered properly. A piece can remain fully online by design, and still depend on careful presentation: the quality of the screen, the calibration of colour and brightness, the placement of speakers, the choice between a monitor and a projection, the beam of a projector and the darkness it needs, the rhythm of a multi-channel install, the architecture that holds it, the objects or images it converses with, and even the ambient sound and light that shape the atmosphere. Now, there seems to be a more serious sense of display as a medium in its own right.

At the same time, Berninger addresses that there is also “a body of physical artworks that has been developing steadily for years, that are both deeply informed by art history and shaped by the current contemporary technological tools.” Heft was founded to give that work the kind of context and physical presence it deserves, with earlier showings at fairs like Art Basel Miami and Paris Photo confirming that for them: “we saw that work made with digital and computational tools is often most compelling to collect in physical form and when experienced in person, as an object, in space.” A permanent space lets Heft act on that “all year,” and it supports the kind of social density that he feels can create “lasting connections over time.”

Looking at HOFA in London, D’Anna describes their program as “always intertwined,” putting AI-driven works “alongside painting and sculpture in the same exhibitions, letting them talk to each other.” The key, he insists, is to link “materiality, concept and process across mediums rather than presenting digital art as a novelty.” The novelty frame, in his view, is the real enemy, and not the technology itself.

Infrastructure: hardware and long-term care

Once digital work is treated as contemporary art rather than tech novelty, its ‘problems’ start to look less glamorous and more like operations. D’Anna is explicit about the layers of difficulty: “hardware (screens, projectors, sound, servers) needs to be high quality, reliable and aesthetically integrated into the space.” Multi-channel and interactive work means “more complex setups, testing and maintenance.” And then comes the collector-facing side: explaining “file formats, editions, on-chain vs off-chain storage,” plus “clear documentation and long-term support” around installation, updates, and “hardware obsolescence.” What changed and will matter for 2026 is that HOFA launched an app “that allows for streaming across any screens in both residential and hospitality environments.” The question all the curators and gallery owners are thinking about: Where does the artwork go, and how does it get to live?

For Heft, white-labeling Verse’s technology into its website was an important step to create coherence between online presentation and the program’s physical reality. He also describes increased control of physical production – “printing, framing, and custom fabrication” – framed not as standardization but as “responding carefully to each artist’s needs.” I feel this is an important subtext: these avant-garde galleries are also somewhat production studios, and not just sales channels. Computational work is shaped into material forms and the gallery’s value lies in the craft, logistics, and long-term care – the competencies that determine whether a work survives its first five years.

Artificial Intelligence

If there is one area where 2025’s “hype to practice” shift becomes visible in the work itself, it’s AI. D’Anna says he’s excited that “technical innovation and conceptual depth are finally aligned.” He describes AI models trained on “highly personal or carefully constructed datasets, instead of scraping the entire internet.” He mentions immersive works that respond “to sound, environment or audience presence in real time,” blurring “the line between artwork and performance.” On the conceptual side, he points to recurring concerns: “memory, identity and data,” ecology and environment, and “new forms of authorship,” where AI is not the artist’s replacement but a collaborator “that must be critically directed, just like a brush.”

Nguyen mentions she’s particularly interested in “AI slop,” and in how quickly AI has been “embraced by a mass audience, including younger generations.” She hopes that once the novelty wears off, “a more critical social consciousness” will emerge, “and that this might mark the beginning of a deeper and more sustained appreciation of digital art as an artistic practice.” Zooming in on the point she makes, the cultural flood of low-grade AI imagery might be the thing that forces a sharper public distinction between practice and noise. Not because the public becomes more sophisticated overnight, but because they become tired, and boredom can be educational.

She also voices a worry that feels relevant beyond art: “the longer-term mental, cognitive, and intellectual consequences” of tools like ChatGPT or Gemini – consequences “we will only fully understand in a few years.” That anxiety tends to be dismissed as reactionary pessimism about new tech, but I feel Nguyen frames it as an honest truth that will shape how digital art is perceived eventually. The medium will be judged not only for what it can do, but for what it does to attention and how it makes you feel.

The blockchain

From AI to the blockchain. Both D’Anna and Nguyen describe the blockchain as an infrastructural technology drifting toward invisibility. Nguyen notes that blockchain is already being adopted at scale in finance. Once it becomes “largely invisible,” crossing “from novel innovation to infrastructural technology,” it may follow a trajectory similar to mainstream AI: omnipresent, mundane, hard to opt out of. For digital art, she hopes that shift will clarify blockchain’s “specific affordances (like blockchain as a medium) and limitations as an artistic form.”

D’Anna’s perspective is more pragmatic and market-facing: he calls out “speculative NFT projects with little artistic substance” as still over-hyped, while arguing that under-explored territory includes “long-form generative works that evolve with time, location or data,” plus “new economic models such as streaming, subscription and on-chain royalties that can provide recurring income for artists and fairer participation for communities.” Royalties to him is an economic instrument, part of an ecosystem that needs to support artists beyond a single sale.

Santoro, meanwhile, describes Vertu’s shift “toward on-chain work and a deeper exploration of what it means to operate within those constraints.” An interesting observation he made: a traditional print background translates well to blockchain-native work because of “compression, limited color palettes, and formal restrictions.” He links those constraints to “silkscreen, plotter drawing, and printmaking,” calling the pairing “especially elegant” when artists are “already building within constraint based systems.”

He names one artist as a key reference: Diid, whose approach to smart contracts and historical contextualization Vertu has been committed to “for several years.” Santoro emphasizes Diid’s “historical understanding of what makes art compelling” and his ability to reference “what came before him.” He also points out “the tools he builds for the community,” which he feels matters because it frames blockchain practice not only as aesthetic production but as infrastructural authorship. This is where 2026 may become more interesting than 2025: not in bigger spectacles, but in deeper literacy about what the medium can structurally do.

What about the collectors?

All four describe some version of collector maturation, but none present it as the victory of 2025. D’Anna says web3-native collectors “remain extremely important,” but their behavior has matured: “less focus on rapid flipping, more interest in provenance, narratives and long-term relationships.” He also notes that collectors are “embracing more the digital medium that also has physical representation,” which is both an aesthetic preference and a practical one: many collectors want a work that can live with them in a legible way.

Berninger describes “meaningful growth” across web3-native and traditional communities, with a recurring emotional detail: the “visible moment of surprise” when collectors encounter work made with contemporary technologies that feels “deeply human, rigorous, and emotionally resonant.” The implication is that a certain kind of collector skepticism dissolves through this embodied experience.

Nguyen hopes “to see more collectors genuinely interested in computer art as a field, rather than approaching it through hype cycles.” She notes a growing group coming from tech who “intuitively understand this work,” but whose buying can remain “emotionally driven” or “strongly influenced by market signals.” That’s why she advocates for more advisers, curators, and art historians to contextualise the field. “Ultimately, time is essential,” she says, because the ecosystem learns through cycles of “public enthusiasm, (over)correction, and readjustment. The more cycles we observe, the wiser we become – and I include myself in that process, as I am still learning and actively seeking insight from those who have been engaged with these practices for much longer.”

Santoro argues that misunderstandings persist because “the way the work is presented has not meaningfully evolved yet.” He points to the abstract conceptual demands of blockchain-linked works: if someone isn’t “fundamentally grounded in this world,” engaging can be challenging. Still, he says the friction often leads to “fun and exciting” conversations – people sense novelty, “that recognition that something new is happening,” and respond with curiosity. Over time, that newness will become familiar, but it will take “repetition, continued exposure, and ongoing education.”

Context becomes the product

Looking at the notes of the interviews, it seems that the medium is not the only thing that needs to mature; the language around it does too. Berninger talks about expanding the role of the gallery as a “social and intellectual space,” with greater emphasis on “contextualization through writing, conversations, and institutional engagement.” Events, gatherings, informal exchanges – the things that often don’t show up in sales reports – are part of the infrastructure and are needed for the work to be taken seriously.

D’Anna describes HOFA’s Digital Art Awards as a “serious, curated framework for recognising excellence in digital, generative and AI-driven practices – rather than a ‘crypto trend’ moment.” Awards, in that context, are not only PR; they are a curatorial device. HOFA integrated finalists into physical shows and brought them into conversation with “more established artists,” using the Awards as a bridge between “digital-native creators and our traditional collector base.” The first initiative of 2025 drew “over 250 applications” and is set to launch its second edition in May 2026.

For the Awards there were commissions across categories, with winners including Maja Petrić, Zach Lieberman, Deekay, and Operator. The message I’m getting here is that ‘digital’ is definitely not one look, but a set of ideas, methods, styles, and contexts. D’Anna also points to HOFA’s Encounters program with Scorpios and Hoho House, connecting artists with music, hospitality, architecture, and technology, as a reminder that digital work is increasingly situated in lived environments.

Lessons offered

When asked for a single insight, D’Anna’s advice to artists is direct: “be very clear about why you are using a particular technology. The medium should serve the concept, not the other way around,” and artists should consider “how your work will live technically in 10–20 years.” To curators: build committed relationships and learn the technical side so you can advocate for “proper installation, conservation and context.” To collectors: “Look at the ecosystem around the artist: their gallery, their long-term vision, how their work is supported and presented.”

Berninger’s advice: “There is an overwhelming amount of art in circulation today, and a lot of noise around it.” He urges people to spend time with work that “truly moves you,” to be honest about what gives it “depth and rigor,” and to set that as the bar – for what you make, show, and invest in.

Nguyen talks about pacing, looking for long-term commitment over quick wins, patronage over short-term logic, and collaboration between younger collectors and institutions. And Santoro’s advice is embedded in repetition: “we have to take care of continued exposure, education, and the patient building of familiarity so that artists aren’t constantly reintroduced as novelties.”

Looking ahead to 2026

If I were to try to compress these conversations into one shared outlook, it might be this: 2026 will not be defined by a single platform, a single hype wave, or a single aesthetic. It will be defined by infrastructure – conceptual and practical, by how many parts of the art world are willing to do the slow work, and by the other parts recognizing and appreciating this.

Positively, none of the four sound like they’re bracing for collapse. They sound like they are building for endurance. D’Anna describes 2025 as the year digital art “started to mature into infrastructure and culture, not just a market cycle.” Berninger describes a lineage – from systems art in the 1960s and 70s to today’s networked condition – that makes this moment feel less like a rupture and more like a new chapter in a longer argument. Nguyen keeps insisting on the long game, the slow alignment between research and market. Santoro describes the field’s “speedrun” of recognition while still emphasizing how long it takes for understanding to settle.

Which is to say: 2026 may not deliver a headline, but it may deliver the normalization of seriousness. The work now seems to be met with the kinds of questions art needs in order to survive: not is it digital?, but what is it doing, who is it for, why is it relevant, and how can we sustain this?