AcademiaIdentity

On being used as evidence: institutional branding in the age of AI

I asked Google's Gemini to generate a profile of me based on everything it could find about me online. I know what you might think. I wasted water for a frivolous experiment. But after a friend of mine told me about the niche sources of information AI could find about him, I was curious. Above all, I was shocked by the level of detail with which it had assembled a picture of my professional and intellectual life. Of course, some things were incomplete, and certain connections were perhaps more speculative than factual. But, all things considered, the profile was remarkably comprehensive.

It was also a little unsettling to see how much of a person can be pieced together from publicly available information, and how easily different activities, affiliations and interventions can be turned into a coherent narrative. It made me laugh (actually, it made me want to cry a little) to notice how Gemini seemed to be calling me out on my own political inconsistencies. Referring to my criticism of institutional approaches to DEI and "decolonization," it noted - with a tone I was certainly not imagining! - that I had argued that one of the most effective ways of neutralizing an antiracist movement was to institutionalize them by establishing committees and dividing activists by offering them a little money and symbolic access to institutional power. It recalled an interview following the publication of my essay Een draaiboek voor dekolonisatie: hoe het niet moet (A Playbook for Decolonisation: How Not to Do It), in which I criticized what I considered to be superficial branding exercises undertaken by institutions such as universities, museums, and cultural centers claiming to "decolonize."

Gemini then proceeded to juxtapose these criticisms with my own institutional engagements. It mentioned my participation in a 2019 jury for the Flemish Audiovisual Fund (VAF) - which I had personally forgotten about already - as well as my supposed involvement in the selection committee for Muntpunt library's decolonization programme, alongside other semi-public antiracist figures. My AI-generated profile presented these activities as evidence of a certain tension that questions how someone so critical of institutions could also participate in their decision-making structures and initiatives.

It is a fair question, I guess? I had obtained a PhD on a research funding scheme (yes, it even included the exact grant number), I had served on a jury for a student film award (for which the compensation was negligible, by the way), I suggested a number of book titles to a friend who was temporarily appointed as a curator at a public library. The library subsequently acquired some of these books (I was not paid for this contribution, nor did it lead to any sustained collaboration or further engagement with the institution). This is to say: besides my PhD, these were relatively minor, informal and, in themselves, unremarkable activities. What interests me, however, is what happens to these interactions once they enter a publicized institutional archive.

My name, alongside those of other artists, academics and activists of color, currently appear in annual reports, funding applications, policy documents and institutional communications. These documents are not just administrative records, but instruments of institutional self-presentation. They provide evidence that an institution is connected to particular communities ad topics, that it has consulted the right people, that it has opened itself up to different perspectives, which may also demonstrate a commitment to antiracist work. Most importantly, this evidence usually has material consequences. It can strengthen funding applications, legitimize institutional programs and help secure the resources that allow these organizations to continue operating, even - as trends will die out - without the names that gave them the credential evidence in the first place. In other words: names - including mine - become resources.

This is precisely the kind of dynamic I was concerned with in Een draaiboek voor dekolonisatie. The institutional response to demands for structural transformation often involves the creation of initiatives that function as mechanisms of containment. Activists of color are invited inside, their expertise is somewhat acknowledged, their presence is documented, and their names become part of an institutional narrative of progress. Meanwhile, the underlying distribution of power, resources and decision-making authority remains largely untouched. The issue, then, is not simply whether institutions include "diverse communities," but what they do with that inclusion, and what their efforts of inclusion allows them to claim.

I would argue that the institutional use of people's names is one of the less examined dimensions of the phenomenon Cedric Johnson (2022) referred to as "gestural antiracism". It is a particularly efficient form of symbolic inclusion because it does not even require the sustained presence of the person. Sometimes all an institution needs is a name. Perhaps a brief interaction or an association that can be translated into evidence of engagement. But the person's name alone becomes a signifier of benevolence, while the institution retains control over the meaning and value of that signifier.

And I am certainly not the only person to have experienced this. People of color working in the cultural sector are routinely mentioned in funding applications, project descriptions or institutional reports without their knowledge, or without any substantial collaborative proposal having been made. Sometimes the relationship is real but far more limited than the institutional account suggests. Sometimes it consists of a single conversation, a recommendation or an informal exchange. The opposite is also true. Without your name being out there, you are technically worthless, no matter the quality of your work. Institutions thrive on perceived proximity to mediocre big-names, while rendering visionary (read: disrupting) small-names invisible. There's nothing new here, but the age of AI did bring about new dimensions to this formula.

Gemini assembled my professional history by drawing on publicly available information, including interviews, essays, institutional documents and references to my participation in various initiatives. It then turned these fragments into a coherent profile, complete with an interpretation of my activism, political critique and its apparent contradictions. In doing so, it inadvertently reproduced something that institutions themselves have been doing. It transformed scattered instances of participation into a narrative of legible institutional engagement. The difference is that Gemini made the process of (reputational) curation visible. It placed my criticism of institutional co-optation alongside the very institutional activities that could be used to complicate, qualify or even neutralize that criticism.

The irony is that AI systems increasingly rely on the digital traces of our professional and public lives to construct (political) profiles and assess expertise. These traces are produced within existing institutional and cultural economies of visibility. If institutions have already converted someone's presence into evidence of their own diversity credentials, AI can further circulate and consolidate that representation. It's not about whether AI gets the facts right, but what happens when the categories, narratives and institutional claims embedded in publicly available data are reproduced at scale, stripped of context and presented as objective knowledge over and over and over again. Who will be the heroes of the future?