AI system
- System
-
The Klimt Color Enigma colourisation model (Google Arts & Culture Lab)
- Model
-
Not assessed
- Release
-
Not assessed
- Application developer
-
Google (Emil Wallner) - Role
-
Not assessed
- Configuration
-
Not assessed
- Inputs and outputs
-
Not assessed
- Autonomy
-
Not assessed
- Workflow
-
Surviving black-and-white photographs of the destroyed paintings. A colorization model pre-trained on 91,749 artworks and fine-tuned on Klimt's color paintings. Curator-compiled written color testimony injected as fixed colors.
- Human oversight
-
Not assessed
- Relation to AI
-
The model proposed colors for the monochrome photographs. Known colors from Smola's archival research were added directly, so the output mixes documented evidence with learned statistical guesses. The Belvedere calls it an attempt at reconstruction, and no source claims the colors are verified.
- AI contribution
-
The model proposed colors for the monochrome photographs. Known colors from Smola's archival research were added directly, so the output mixes documented evidence with learned statistical guesses. The Belvedere calls it an attempt at reconstruction, and no source claims the colors are verified.
Benefits
- Outcome identifiers
-
Not assessed
- Benefit
-
ML color reconstruction of Klimt's destroyed Faculty Paintings launched 7 October 2021 with the Belvedere, shown in its 2025 exhibition
- Category
-
Increased Creativity
- Status
-
Not assessed
- Beneficiary groups
-
Organizations. Society.
- Beneficiaries
-
Not assessed
- Reach
-
Not assessed
- Distribution and access
-
Not assessed
- Duration
-
Not assessed
- Expected?
-
Yes
- Costs and harms
-
Not assessed
Measurement
- Studies
-
Not assessed
- Study design
-
Not assessed
- Measurement
-
Not assessed
- Setting and sample
-
Not assessed
- Measures
-
Not assessed
- Comparison
-
Assumed alternative: the surviving black-and-white photographs of the destroyed paintings
- Results
-
Assumed comparison. Reference: have only survived in grainy black-and-white photographs, devoid of the sumptuous, aurous pigment on which the painter's name was made. Result: a digital reconstruction of his Faculty works — in full and glorious color.
- Additional value from AI
-
Not assessed
- Limitations
-
Not assessed
- Denominators
-
Not assessed
- Materials for replication
-
Not assessed
Money
- Kind of value
-
Not assessed
- Amount
-
Not assessed
- Currency and prices
-
Not assessed
- Period and unit
-
Not assessed
- Realization
-
Not assessed
- Recipients and payers
-
Not assessed
- Gross or net
-
Not assessed
- Included costs
-
Not assessed
- Comparison
-
Not assessed
- Method
-
Not assessed
- Return and payback
-
Not assessed
People and places
- Participants
-
Developer:
Google (Emil Wallner). Operator: Belvedere, Vienna, with Google Arts & Culture. - Jurisdictions
-
Benefit: Vienna,
Austria. Developer: 
France. Operator: Austria. - Ownership and control
-
Not assessed
- Dependencies
-
Not assessed
- Cross-border connections
-
Not assessed
- Geopolitical implications
-
Not assessed
Follow-up
- Conditions for replication
-
Monochrome photographs of adequate quality exist. Written color descriptions exist for key areas. An institutional partner validates and displays the result.
- Replication attempts
-
1 follow-up entries. Benefit recurred: 1.
- Later findings
-
Not assessed
- Availability
-
Not assessed
- Institutional decision
-
Not assessed
- Grounds
-
Not assessed
- Action and result
-
Not assessed
- Success criterion
-
Not assessed
Evidence
- Source numbers
-
AIBD-0074-source-0001–0004
- Citations
-
Numbered references below.
- Language
-
Not assessed
- Source relationships
-
Operator, Independent, Developer
- Original evidence
-
Not assessed
- Access and versions
-
Not assessed
- Supported findings
-
Not assessed
- Evidence links
-
Not assessed
- Evidence strength
-
Not assessed
- Missing evidence
-
Not assessed
Grades
- Assessed benefits
-
Organizations. Society.
- Depth
-
Organizations: 1. Society: 1.
- Reasons
-
Organizations: Small: the Belvedere adopted and exhibited the reconstructions in 2025, per its exhibition page. Society: Small: the public can view a color hypothesis of lost works; Smola himself called the result surprising where no color knowledge existed (Jing Daily).
- Observed or intended
-
Not assessed
- Rubric
-
Not assessed
- Assessor and date
-
Not assessed
- Alternative assessments
-
Not assessed
- Fill
-
Independently reported
Entry details
- Kind
-
Not assessed
- Related entries
-
Same task, AI reimagining a canonical painting shown to the public: Google surfaced an AI-generated version of Vermeer's “Girl With a Pearl Earring” as its top result, misleading users, the public the Klimt reconstruction also reaches.
- Field
-
Art. Museums and art history.
- Purpose
-
Colorize black-and-white photographs of destroyed paintings in the artist's style, guided by archival color descriptions
- Dates
-
Date basis unassessed: 2021 (precision unassessed). Updated: 22 October 2026.
- Stage
-
Not assessed
- Version
-
SHA-256: 96c3a11a391b…
References
- Other identifiers
-
AI Incident Database: 554.
- Category mappings
-
Not assessed
- Brief
-
Not assessed
- Revisions
-
8 revisions. Latest: 22 October 2026.
Sources
-
News report Independent author
Chen, M. (2021). Google Arts & Culture And Belvedere Museum Trained An AI To Color Like Klimt. Jing Daily Culture, 7 October 2021.
on October 7, Google Arts & Culture unveiled Klimt vs. Klimt – The Man of Contradictions
-
News report Independent author
I Programmer (2022). Google AI Recreates Lost Klimt Artworks. i-programmer.info.
was trained on 91749 artworks from Google Arts & Culture to enable it to learn object boundaries, textures, and frequent compositions in artworks
-
Press release Deployer
Belvedere (2025). Gustav Klimt – Pigment & Pixel, exhibition page, 20 February – 7 September 2025. belvedere.at.
Now for the first time, the Ceiling Paintings can be viewed in color in their original size and arrangement on the ceiling.
-
Press release Developer
Google Arts & Culture (2021). The Klimt Color Enigma. artsandculture.google.com.
developed an algorithm to use Dr. Smola's research to restore the Faculty Paintings