Jonathan Johnson-Swagel

Judith Butler’s theory of performativity redefined the understanding of identity as a construct, not an essence. According to Butler, gender is not something one is but something one does, an iterative act produced and regulated by norms. These acts, repeated over time, solidify into what appears to be natural or fixed. Butler’s framework, initially rooted in gender, provides a lens to interrogate how algorithmic systems now perpetuate, enforce, and even produce new identities. In the age of AI, the performativity of identity shifts, expanding beyond the individual to a realm mediated by algorithmic surveillance and prediction. This explores the interaction between performativity, surveillance, and resistance, offering an analysis of how identity itself becomes algorithmically disciplined and remade.

The Algorithmic Gaze: Surveillance as Identity Formation

Foucault’s notion of the panopticon, as discussed in the previous article, intersects here with Butler’s performativity. Under the algorithmic gaze, individuals adjust their behaviors in response to systems that predict and evaluate their actions. The performative acts Butler identifies—those repetitions that constitute gender or identity—are increasingly shaped by the invisible algorithms that observe and reward conformity. The power of these systems lies in their ability to operate imperceptibly, producing subjects who internalize algorithmic expectations even without conscious recognition.

Consider the example of social media platforms. These platforms function as digital panopticons, where algorithms amplify posts, images, and behaviors that align with their calculated understanding of engagement. For women and marginalized groups, this often means conforming to narrow standards of beauty and desirability that reflect the biases encoded in the system’s training data. Beauty filters, for instance, actively impose Eurocentric and heteronormative ideals, subtly disciplining users to align their self-presentation with algorithmic norms. In such cases, identity is not merely performed for an audience of peers but for the algorithms themselves, which determine visibility and social validation.

This shift marks a departure from traditional performativity: whereas Butler’s performative acts are regulated by cultural norms, algorithmic performativity operates at a meta-level. Norms are encoded into algorithms, which then reward or punish behaviors, creating a feedback loop where the self is disciplined not by societal watchers but by systems designed to quantify engagement and conformity.

Gender and the Codification of Norms in AI

Butler’s analysis of gender as a performance that sustains power structures resonates deeply in the context of AI. Algorithms, trained on historical data, are inherently conservative: they reproduce the biases and inequalities embedded in their datasets. Gender, as Butler argues, is constructed through norms that are policed and maintained through repetition. AI systems amplify these repetitions, reinforcing gender binaries and stereotypes with extraordinary efficiency.

Take virtual assistants like Alexa or Siri as case studies. These AI systems are gendered by design, given feminized voices, and coded to be subservient and polite. Their roles as helpers reflect and perpetuate the cultural expectation that women are caregivers and service providers. The repetition of these gendered performances, now scaled to billions of interactions, reifies outdated and oppressive norms.

More insidiously, algorithms enforce conformity by rewarding adherence to normative expectations. Consider TikTok’s recommendation system, which privileges “algorithm-friendly” content—posts that align with trends and dominant cultural aesthetics. Creators whose appearances or behaviors deviate from these norms are algorithmically penalized, receiving less visibility and engagement. The result is an ecosystem where identity is disciplined into narrow, quantifiable categories, erasing fluidity and nonconformity.

The Feedback Loop of Algorithmic Performativity

Under algorithmic performativity, individuals are not just performers but data points. Their actions, appearances, and preferences are continuously quantified and fed back into systems that refine and reinforce norms. This feedback loop creates a form of performative self-discipline that exceeds even Butler’s framework.

A person scrolling through Instagram might unconsciously tailor their posts to gain more likes, aligning their self-presentation with algorithmic trends. Over time, this behavior solidifies into a habit, a new mode of self. The individual becomes complicit in their own discipline, performing not just for human peers but for a machine that governs visibility.

What distinguishes this from traditional performativity is the scale and speed of enforcement. Norms are no longer policed by cultural institutions alone; they are encoded into systems that operate globally, invisibly, and instantaneously. The algorithmic feedback loop collapses the gap between performance and policing, creating a world where identity is both performed and policed in real time.

Resistance Within and Beyond the Algorithm

Despite the totalizing nature of algorithmic systems, Butler’s concept of resistance remains relevant. Performativity is not a closed system; it always contains the potential for subversion. The repetitive nature of performative acts means they can be disrupted, reconfigured, and repurposed.

In the context of AI, resistance can take many forms. Artists and activists have used adversarial attacks to disrupt facial recognition systems, challenging the biases that these systems reproduce. For example, wearing makeup or accessories that confuse facial recognition cameras becomes a performative act of resistance, disrupting the algorithmic gaze.

Additionally, generative AI offers possibilities for reimagining identity outside of normative constraints. For example, creators can use AI to design avatars or narratives that explore fluid or hybrid identities, challenging binary representations of gender and race. These acts, though limited, suggest pathways for leveraging the very systems of control to generate new modes of being.

However, resistance must also address the structural foundations of algorithmic power. Policymakers, technologists, and communities must work together to expose and dismantle the biases encoded into systems. Butler’s work reminds us that identity is not fixed; by refusing to repeat the scripts handed down by algorithms, individuals and societies can disrupt the norms that seek to discipline them.

Toward a New Ethics of Performativity

As we move deeper into the algorithmic age, the ethical implications of performativity demand attention. If algorithms now shape the identities they observe, what responsibilities do designers and users bear for the outcomes? Butler’s emphasis on the constructed nature of identity invites a radical rethinking of how AI systems are designed. Rather than reproducing oppressive norms, these systems could be programmed to challenge them, fostering pluralism and creativity.

This vision requires a shift in how we conceive of AI—not as a neutral tool but as an active participant in the construction of identity. By embedding principles of diversity and fluidity into AI systems, we might create technologies that expand, rather than constrain, the possibilities of who we can become.

Conclusion

In the algorithmic age, Butler’s theory of performativity becomes a framework for understanding not just the construction of identity but its algorithmic mediation. The interaction between individual performance and algorithmic discipline produces a new form of subjectivity: one that is shaped by systems designed to observe, predict, and reward conformity. Yet, as Butler reminds us, performance always contains the seeds of its own undoing. Through acts of resistance, both individual and collective, the scripts imposed by algorithms can be rewritten.

The question that remains is whether we can harness these systems to expand identity’s horizons or whether we are destined to remain disciplined by the invisible gaze of the machine. The answer, perhaps, lies not in rejecting the algorithmic but in reimagining it, transforming the mechanisms of surveillance into tools for liberation.