In Himalayan sacred art, the artist’s name was never meant to be part of the work. This is especially true of Paubha, Thangka paintings and statues from Nepal, where scholars have identified only a handful of named traditional artists across several centuries.
In the traditional art community, painting sacred figures was understood not as a pursuit of personal recognition, but as an act of dharma and offering. The work carried spiritual purpose, where devotion, discipline, and service to the tradition mattered more than individual authorship. This anonymity was intentional and consistently practiced across centuries.
Today, however, that same anonymity creates a serious challenge for those working to protect and document the tradition.
Understanding The Market
Thangka and Paubha are living artistic traditions. Artists continuously create new works for devotional, ceremonial, and community purposes. A newly completed Thangka is a legitimate sacred object, created in the traditional way. It serves the same purposes as works painted centuries before it. Age has never been the measure of authenticity in this context.
The Thangka and Paubha market has always carried a natural range of works. A master artist who spends months on a single painting, grinding mineral pigments by hand and applying thousands of precise strokes, commands a high price. That price reflects time, skill, and decades of training. A less experienced artist working in the same tradition, using more accessible materials and completing a work more quickly, prices their work accordingly. Both are genuine. Both serve the tradition. Price reflects the depth of skill and time invested, not a hierarchy of legitimacy.
At the far end of that range sit printed souvenir items. Machine-printed images of deities have long served a separate purpose. They are affordable objects with diverse purposes but cannot commission or purchase handmade value. Most buyers understand what they are purchasing. These items exist in their own category and serve a real need.
Now the problem is not the range but misrepresentation.
The first form of misrepresentation is antiquity fraud. A recently painted or printed work enters an antique auction or goes to a collector as a historical artifact, centuries-old. In this context, the new painting is fraudulent not because it is new, but because its seller claims it is old. These kinds of issues can be solved by material dating tools like carbon-14 analysis and XRF.
The second form of misrepresentation is quality fraud. A machine-printed and retouched copy sold as a handmade traditional work deceives the buyer about how the work was made. The buyer pays for skilled handiwork and receives something else entirely. This directly undercuts artists who invest years in genuine training.
The third form of misrepresentation is attribution fraud. A work sold under the name of a recognised living master when it did not come from that master’s hand. This is the most complex problem in the contemporary market. In Thangka and Paubha practice, artists work within family lineages and workshop traditions. Students paint closely in their teacher’s style. This is intentional. It is how the tradition transmits itself, and it carries no deception within the community where everyone knows the relationships involved.
But in an international art market, a collector cannot visit a workshop or verify a lineage directly. A work painted by a master’s student and a work painted by the master personally may look identical to an outside buyer. Without documentation created at the time the work is painted, establishing correct attribution after the fact becomes very difficult.
This is the gap that digital documentation directly addresses. The Himalayan Art Council established its provenance system to create verified records for living artists at the point of creation, before works enter a market where their origin can be misrepresented or disputed.
The Limits of Material Analysis
Each form of misrepresentation calls for different tools. The scientific methods developed for dating older works address antiquity fraud effectively. For attribution fraud in a living tradition, they reach their limits quickly.
Carbon Dating
Carbon-14 dating helps establish when materials in a painting were made, and researchers have used it to study early Nepali paintings in major international collections. For the living Thangka and Paubha tradition, it answers the wrong question entirely. Dating the canvas of a work painted last year by a master artist tells a buyer nothing about who painted it or whether the attribution is correct. Carbon dating is a tool for historical archaeology, not for the attribution challenges the contemporary market faces.
X-Ray Fluorescence
XRF scanning maps the elemental composition of pigments using non-destructive X-rays. It serves well for one specific purpose: identifying works that use modern synthetic pigments inconsistent with traditional practice. If a painting sold as traditional handmade work contains titanium dioxide, which manufacturers only began synthesising for paint use after 1921, or other modern synthetic compounds, XRF exposes the inconsistency immediately.
Beyond that specific use, XRF reaches its limits quickly. The technology identifies elemental composition, not specific compounds. A reading confirms copper is present but cannot tell whether that copper came from authentic malachite or a synthetic substitute. Paint layers also generate overlapping signals, so readings from the original work, subsequent touch-ups, and ground layers blend together.
Most critically, a work painted with traditional mineral pigments and hide glue binder, whether by a recognised master or by someone misrepresenting their work, passes every XRF scan. The chemistry reads as consistent with traditional practice because the materials are consistent with traditional practice. High-specification MA-XRF systems cost upward of $200,000, and even portable XRF units at lower price points remain largely inaccessible to most local authenticators and smaller institutions in Nepal.
Binder Chemistry
Traditional Thangka paintings use animal hide glue as both canvas primer and paint binder. Researchers use Pyrolysis-Gas Chromatography-Mass Spectrometry (Py-GC-MS) to analyse microscopic paint samples and identify binder composition and age markers. Research on Himalayan and Chinese Buddhist architectural paintings suggests these methods can distinguish ancient animal glue binders from fresh ones.
For antiquity fraud, this has clear relevance: a work claimed to be centuries old but painted recently carries a binder that has not undergone the expected protein degradation. For attribution questions between living artists working in the same tradition, however, binder chemistry reaches the same wall as the other material methods. All traditional artists use the same hide glue. Chemistry cannot tell one painter’s hand from another’s.
One further consideration: Py-GC-MS requires a small physical sample from the painting. The sampling is minimal but not entirely non-invasive. For a consecrated sacred object, practitioners and institutions need to weigh that carefully before proceeding.
Where AI May Add Value
Brushstroke Analysis
A master painter works from decades of accumulated muscle memory. The brush moves without conscious direction. Someone copying that work must look at the original, interpret it, and then guide their hand. That additional cognitive step introduces microscopic inconsistencies invisible to the naked eye but potentially measurable in high-resolution digital scans.
Published research from European art supports this approach. In 2017, a team at Rutgers University led by Professor Ahmed Elgammal analyzed approximately 300 digitized drawings by Picasso, Matisse, Schiele, and Modigliani comprising over 80,000 individual strokes. Using deep neural networks, the system classified individual strokes with 70 to 90 percent accuracy and attributed entire drawings with above 80 percent accuracy. In every test, the system identified commissioned forgeries. “A human cannot do that,”- Elgammal noted.
This research used European line drawings where individual strokes are clearly visible. Thangka painting, with its precise outlines of sacred figures painted through trained brushwork, presents a plausible candidate for a similar approach. However, researchers have not yet published validation of brushstroke analysis on Himalayan sacred art at scale. This remains a direction worth pursuing rather than an established technique.
Iconometric Analysis
Every Thangka deity follows strict iconometric guidelines: precise mathematical ratios governing a figure’s eyes, torso, posture, and limbs. Canonical texts set these proportions and practitioners follow them closely. At the same time, decades of painting the same deity may develop subtle and consistent personal habits in how individual masters interpret those proportions in practice.
AI keypoint extraction can map coordinate points across a deity’s figure, calculate ratio matrices, and compare them against verified works by a known master. In principle, someone copying the composition replicates the general proportions but may not capture the specific mathematical texture of an individual master’s long-established habits.
Researchers have not yet published validation of this approach on Thangka painting specifically. The underlying logic draws on established methods in computational art analysis, but institutions exploring these tools should treat the results as supporting evidence rather than standalone proof.
What No Technology Resolves
Three limitations apply across all attribution methods.
The Data Problem
Brushstroke and iconometric analysis depend on verified reference databases of a known master’s confirmed works. For well-documented masters whose work major institutional collections hold, researchers are building these databases. For a painter from a smaller regional or family workshop, the reference set may contain very few documented works or none at all. Without sufficient reference material, the analytical tools cannot deliver reliable results.
The Lineage Problem
Traditional Thangka and Paubha training explicitly develops work that aligns closely with the teacher’s style. This is the tradition’s method of transmission, and it is intentional. An apprentice typically spends ten to fifteen years learning under one master. A student who trained under a specific master for many years may paint work that proves genuinely difficult to distinguish from that master’s through brushstroke or iconometric analysis alone. The tools cannot resolve what is, in some respects, a deliberate feature of how the tradition sustains itself. This matters considerably for market valuation, even though the closeness of style carries no negative implication within the tradition itself.
The Materials Problem
A work painted with traditional mineral pigments and animal hide glue binder will pass material analysis regardless of who painted it. No analytical method currently distinguishes one trained artist’s materials from another’s, because the materials are the same. Attribution in this tradition ultimately depends on documentation, testimony, and the judgement of people with deep knowledge of individual artists’ work, not on chemistry alone.
A Question Worth Holding
Thangka painting developed as an anonymous practice. Painters understood their identity as irrelevant to the painting’s purpose and devotional function.
Building authentication systems and provenance records, even with genuine care for the tradition, changes something about how these objects move through the world. It reframes them, at least partially, as attributable and certifiable assets. Whether that shift serves or complicates the living tradition is a genuine question. It belongs to the communities of practitioners, scholars, and devotees who have sustained this art across centuries. Institutions working with this material are right to hold that question openly as their tools develop.
The Current Picture
AI is not saving Thangka art. It gives authenticators better investigative tools, creates additional obstacles for sophisticated forgeries and misrepresentation. It offers artists stronger grounds to expect that their work receives the recognition it deserves in international markets.
These tools work best alongside expert judgement. Brushstroke data supports what trained observation suspects. Iconometric comparison tests what expertise intuits. Neither is sufficient independently. Together, a specialist with deep subject knowledge interprets them to raise the practical cost of misrepresentation meaningfully.
The tradition of unsigned work remains. What changes, incrementally, is whether the knowledge and devotion behind that work receives the recognition it has always deserved.
The Himalayan Art Council documents and certifies traditional Himalayan artworks through a digital registry, assigning each work a tamper-proof digital certificate recording authorship, and ownership history.