Take a JPEG image, open it in a text editor, delete a few lines of code, and save it. When you reopen the image, it will look broken. Colors shift. Bands of static replace sections of the image. The compression artifacts that the JPEG format normally hides become visible as pixelated blocks and color streaks. The image is corrupted, and in the conventional understanding of digital media, corruption is failure. But in glitch art, the corruption is the point. The error is not a mistake to be fixed. It is the material the artist works with.
Glitch art is the practice of using digital errors, system failures, and data corruption as aesthetic and conceptual material. The term "glitch" comes from electronics and computing, where it refers to a brief, unexpected malfunction in a system. Glitch artists deliberately introduce errors into digital files, manipulate compression algorithms, exploit hardware failures, or work with the artifacts that digital systems produce when they break. The results are images, videos, and audio works that display the visual or sonic signatures of digital failure: pixel sorting, color channel displacement, datamoshing, compression artifacts, and signal noise.
This entry covers the history of glitch art, the techniques artists use, the theorists who defined the field, and how glitch aesthetics have evolved in the era of AI-generated images.
What Is Glitch Art and How Does It Work?
Glitch art operates on the material structure of digital files. A digital image is not a picture in the way a painting is. It is a file containing data that a computer interprets as an image. The data includes information about pixel colors, compression instructions, metadata, and file structure. When this data is altered, the image changes. If the alteration is intentional and controlled, the result is glitch art.
Several techniques are common in glitch art practice. Databending involves editing a digital image or video file in software designed for a different file type, such as opening a JPEG in an audio editor and applying audio effects, then saving it back as an image. The audio effects alter the image data in ways that produce visual distortions. Datamoshing involves manipulating video compression, specifically the motion vectors in interframe compression codecs like MPEG, to make one video's motion data bleed into another video's image data, producing surreal morphing effects where figures and landscapes melt into each other. Pixel sorting uses algorithms to rearrange pixels in an image based on their values, sorting rows or columns of pixels by brightness, hue, or saturation, producing streaks and bands that follow the image's tonal structure.
Codec manipulation involves directly altering the parameters of video or image codecs to produce errors. A 2026 paper presented at the Expressive Japan 2026 conference proposed "GCNE," a system that presents the internal structures of video codecs as visual modules, allowing artists to construct unique processing flows and structurally intervene in codec algorithms. The system implements compression techniques like DCT (Discrete Cosine Transform) and motion compensation as independent nodes, using GPU acceleration to reflect parameter changes in real time. This represents a technical evolution from the early days of glitch art, when artists hacked files blindly, toward a more structured understanding of how digital systems produce errors. Read the paper at ITE Technical Report.
History and Origins
Glitch art emerged in the late 1990s and early 2000s, growing out of electronic music, hacker culture, and net art. The term "glitch" was applied to visual art by artists and musicians who were experimenting with corrupted digital files, circuit bending (modifying electronic circuits to produce unexpected outputs), and the aesthetics of noise. The early glitch art community was connected to the demoscene (a subculture of programmers creating audiovisual demonstrations on limited hardware) and to electronic musicians like Oval and Yasunao Tone, who used corrupted CDs and modified audio files to produce music.
Rosa Menkman, a Dutch artist and theorist, is the figure most responsible for defining glitch art as a field. Her book Glitch Studies Manifesto (2010) and her Vernacular of File Formats (2010) established the theoretical framework for understanding glitch as an aesthetic and critical practice. Menkman argued that glitches reveal the material structure of digital media, the compression algorithms, encoding standards, and file formats that normally remain invisible. When a glitch occurs, the viewer sees not the image but the system that produces the image. This visibility, Menkman argued, is politically and aesthetically significant because it breaks the illusion of digital transparency.
The glitch aesthetic spread beyond fine art into mainstream visual culture. Music videos, fashion campaigns, advertising, and film title sequences adopted glitch effects as a visual style. By the 2010s, glitch had become a recognizable design trend, with software like Photoshop and After Effects offering built-in glitch filters that simulated the look of data corruption without requiring artists to actually corrupt files. This mainstreaming created a tension within the glitch art community: was glitch still meaningful as a critical practice when it could be produced with a filter?
Key Artists and Masterworks
Rosa Menkman (b. 1983)
Menkman's Vernacular of File Formats (2010) is both an artwork and a theoretical text. It catalogs the visual characteristics of different file format errors, showing how each compression format (JPEG, MPEG, BMP, TIFF) produces a distinct visual signature when corrupted. The work is a taxonomy of digital failure, and it established the vocabulary that glitch artists still use. Menkman's practice combines art-making with critical writing, and her work has been exhibited at the New Museum in New York, Transmediale in Berlin, and other institutions focused on digital culture.
Antonio Roberts (b. 1988)
Roberts, a UK-based artist and curator, creates glitch art that examines the relationship between digital perfection and error. His work uses corrupted data and custom software to produce images and installations that challenge the assumption that digital technology should be seamless. Roberts has curated glitch art exhibitions in the UK and has been an active voice in the community for connecting glitch practice to broader questions about digital culture and ownership.
0xDEAFBEEF (active from 2020)
The anonymous artist 0xDEAFBEEF creates generative works using low-level C code that produce audiovisual pieces with deliberate glitch aesthetics. The work Synth Poems (2021), consisting of 128 unique digital audiovisual works generated by code stored on a blockchain, references early electronic art and analog sound synthesis. At Art Basel 2026, the works were presented as Synth Poem Oscilloscopes, handcrafted interactive sculptures that embodied the digital system in forged iron and electronics. 0xDEAFBEEF's practice connects glitch art to generative art and to the material history of electronic art. Read the interview at hube magazine.
Sarah Rosalena (b. 1982)
Rosalena's 27-foot tapestry Threading the Boundless: Omnidirectional Terrain (2025, LACMA) uses handweaving patterns distorted through computer-generated satellite imagery, producing a visual effect that the artist describes as "almost like it's glitching out." The work is not glitch art in the conventional sense of corrupted files, but it applies the glitch aesthetic, the refusal of digital smoothing, the visible artifacts of computational processing, to a woven medium. This cross-medium application demonstrates how glitch aesthetics have moved beyond their origins in digital file manipulation. Read the interview at The Art Newspaper.
Glitch Art and AI
The rise of AI image generation has given glitch art a new context and new material. AI models produce their own kinds of errors: hallucinations (confident but incorrect outputs), artifacts (visual features that look real but are structurally impossible), and inconsistencies (hands with six fingers, text that is almost readable, architecture that violates physics). These AI errors are the contemporary equivalent of the JPEG compression artifacts that early glitch artists exploited.
A 2025 study published in CINEJ Cinema Journal examined glitch aesthetics in the context of AI, AR, VR, and the metaverse, analyzing how glitch art practices are being reproduced in new technological media. The study found that while the traditional layers of meaning in glitch aesthetics persist in new technologies, they undergo significant transformations. AI-generated visual works produce unexpected artifacts through deep learning models, and these artifacts are being used by artists as critical tools, just as early glitch artists used compression artifacts as critical tools. The study emphasizes that "glitch remains relevant in the digital age both as a critical art practice and an aesthetic strategy." Read the study at CINEJ Cinema Journal.
In May 2026, a talk-lecture titled "Beyond Errors" at the Ruse Art Gallery in Bulgaria explored the concept of error not as failure but as an opportunity for new perception and artistic inquiry. The event addressed the psychological and philosophical dimensions of errors, as well as specific artistic practices related to glitch, including various types of visual distortion, techniques for intentionally "breaking" images, and the transformation of technological malfunction into an artistic gesture. Read about the event at Ruse Art Gallery.
Glitch Art and Critical Practice
Glitch art is not just a visual style. It is a critical practice that questions the assumptions of digital culture. The dominant ideology of digital media is one of transparency and perfection: digital images should be clean, smooth, and indistinguishable from reality. Glitch art breaks this ideology by making the digital visible. When you see a glitched image, you see the file format, the compression algorithm, the encoding standard. You see the infrastructure that produces digital images, and seeing that infrastructure is the first step toward questioning it.
This critical dimension is what distinguishes glitch art from glitch as a design trend. A Photoshop filter that adds pixelated distortion to an advertisement is using glitch as decoration. A glitch artist who corrupts a JPEG to expose the compression algorithm's effect on image quality is using glitch as critique. The visual results may look similar, but the intent and the intellectual framework are different.
Glitch art also raises questions about authorship and control. When an artist corrupts a file, the resulting image is partly determined by the artist's choices (which bytes to delete, which parameters to change) and partly determined by the system's response to the corruption (how the codec interprets the damaged data). This shared authorship between human and machine is a theme that connects glitch art to broader questions about digital art and computational creativity.
Glitch Art Today
Glitch art in 2026 exists in a different context than it did in 2010. The techniques that once required technical knowledge of file formats and compression algorithms are now available as filters and plugins. AI models produce new kinds of errors that early glitch artists could not have anticipated. The critical edge that glitch art once had, its ability to reveal the hidden infrastructure of digital media, has been partially blunted by the mainstreaming of glitch aesthetics in advertising and design.
But the practice continues to evolve. Artists are working with AI hallucinations, codec manipulation tools like the GCNE system, and cross-medium applications that bring glitch aesthetics into weaving, sculpture, and physical computing. The fundamental impulse behind glitch art, the desire to make the invisible visible, to find beauty in failure, and to question the ideology of digital perfection, remains relevant in a world where digital images are increasingly generated by algorithms that most people do not understand.
See Glitch Art in Person
Glitch art is primarily a screen-based medium, but it is best experienced in gallery contexts where the scale, sound, and physical presence of the work can be appreciated. Look for Rosa Menkman's work at digital art festivals like Transmediale in Berlin and at institutions like the New Museum in New York. For contemporary practice, watch the programs of digital art galleries like bitforms in New York and festivals focused on new media art. The intersection of glitch art with NFT art and blockchain-based generative art has also created new venues for viewing and collecting glitch work.
For more on related digital art practices, read our entries on digital art, generative art, and net art, or explore our blog post on how to read a painting to learn more about analyzing visual art across mediums.
