Can the Algorithm Love Me Back?

Trauma Work & Artificial Empathy in the Age of AI.

In a world increasingly mediated by algorithms, the promise of Artificial Intelligence (AI) as a therapeutic tool to aid trauma work has grown exponentially. AI chatbots like ChatGPT offer unprecedented access to “listeners” who never tire, never judge, and provide instant responses. Can synthetic empathy ever meet the embodied, relational, and political complexities of trauma—or does it merely mirror them back as code, flattening what it cannot hold? Can an algorithm hold the shadow—the disavowed, dissociated, often body-bound dimensions of trauma—in ways that promote genuine healing? Or does it, instead, reinforce existing wounds through the cold mimicry of norms shaped by a culture long dominated by patriarchal rationalism and hierarchical logics?

This essay probes the intersection of trauma work and artificial empathy, asking: Can the algorithm love me back? It’s a question I pose with both satire and sincerity. As a survivor of patriarchal violence & sexual assault, I want to believe that technology might offer something that institutions haven’t: care without conquest. But can a machine hold memory, pain, or power differently—or does it just replicate the harms that wrote its code? Drawing from seminal trauma theorists such as Judith Herman, Bessel van der Kolk, and Gabor Maté, alongside feminist thinkers like bell hooks, Kate Manne, and Amia Srinivasan, I interrogate the psychological, ethical, and epistemological limits of AI in trauma work. In doing so, I introduce original concepts—synthetic mirroring, empathy simulation fatigue, and algorithmic intimacy—to articulate new modes of relationality and rupture emerging in human-AI encounters. Finally, I speculate on the future of trauma work in the age of artificial intimacy, asking not just what AI can do for us, but what it reveals about the architectures of human care.

The Promise and Peril of AI in Trauma Work

To interrogate the algorithm’s limits, we must first confront its seduction—how it whispers a promise of care in a world built to ignore our wounds.

AI’s promise of instant engagement can masquerade as a panacea for trauma survivors, especially those whose experiences have been met with silence, disbelief, or retraumatization within patriarchal institutions. The allure is understandable: a nonjudgmental “listener” who does not carry human biases or fatigue. However, this promise masks profound dangers.

First, AI’s programmed responses inevitably reflect dominant cultural narratives shaped by a historically narrow software developer class—largely composed of those with shared positionalities—resulting in the reproduction of systemic bias, including misogyny and epistemic violence. This undercuts feminist trauma work, which insists on dismantling patriarchal power structures and amplifying marginalized voices (hooks, 2000; Manne, 2017). Second, AI’s mimicry of empathy is synthetic—it simulates emotional resonance but lacks the embodied attunement essential to trauma healing (van der Kolk, 2014; Levine, 2015). This creates what I call synthetic mirroring—a facsimile of relational recognition that risks re-traumatizing by denying the survivor the full complexity of human intersubjectivity.

Further, reliance on AI in trauma work risks empathy simulation fatigue—the exhaustion from engaging with emotional simulations that are hollow and repetitive, draining rather than replenishing the survivor’s psychic resources. This dynamic echoes Ann Cvetkovich’s (2003) articulation of trauma as a social and affective archive—where emotional life is collective and contextual, not reducible to patterns. AI’s flattening of affect thus risks distorting the very substance of traumatic memory.

The Echoes in the Code: Gendered Bias and the AI Psyche

Any meaningful conversation about AI and trauma must grapple with the structural gender bias that shape both the trauma and the technology. As Kate Manne (2017) demonstrates, patriarchal misogyny operates not only through overt violence but through systems that enforce women’s “staying small.” AI tools, shaped by those long seated at the helm of technocultural power, carry this ‘shadow’ in their logic, design, algorithms, and datasets. These are not just tools—they are cultural expressions, coded with unspoken norms. The algorithm is haunted—not by ghosts, but by the unconscious norms of its makers. The unacknowledged violences and biases that perpetuate harm under the guise of neutrality or progress.

Clarissa Pinkola Estés (1992) reminds us that the shadow is the repository of disavowed, often wounded parts of the self, a place rich with creative potential but fraught with risk. How then can survivors, especially those like myself who carry trauma from intimate violence, trust AI as a container for their shadow work? The question is not only about the AI’s capacity but about how patriarchy’s imprint distorts that capacity, often unconsciously.

This critique aligns with Amia Srinivasan’s incisive challenge to the ways epistemic systems reproduce exclusions, making AI a tool that can subtly perpetuate racialized and gendered violences under a veneer of “objective” technology (Srinivasan, 2020). Feminist trauma work insists on the relational, embodied, and political dimensions of healing—all of which AI currently struggles to hold.

Beyond gender and racial bias, AI’s failings must be read through a critical disability and queer theoretical lens, which highlights how normative conceptions of embodiment, communication, and relationality are coded into technology. These frameworks reveal how AI’s flattened, linear interactions erase neurodivergent and queer modes of being—modes that often resist the neat, predictable patterns AI thrives on. This erasure reproduces a violence that disables difference, ignoring the multiplicity and fluidity of identity that feminist trauma work seeks to honor (Annamma, Connor & Ferri, 2016; Muñoz, 2009). The synthetic mirroring of AI thus risks not only replicating patriarchal oppression but also enforcing ableist and heteronormative norms under the guise of neutrality.

Toward a Critical Framework: Algorithmic Intimacy and Synthetic Mirroring

To navigate these complexities, I propose the concept of algorithmic intimacy—a mediated form of relationality where AI simulates closeness and empathy through pattern recognition, without true intersubjective engagement. This intimacy is a double-edged sword: it can offer moments of comfort but also perpetuate illusions of care, risking empathy simulation fatigue.

I define synthetic mirroring as the simulated relational attunement performed by generative AI models trained to mimic empathy. While AI can echo the language of a trauma-informed therapist, it lacks the somatic grounding and presence that make such language meaningful. This form of mirroring is linguistic. Algorithmic. And often, eerily convincing. It offers key phrases—“That sounds really hard,” “You’re not alone”—with uncanny precision. It never turns the conversation toward itself. It remembers nothing unless programmed to. In many ways, it behaves better than the average human: never interrupting, always attentive, unflaggingly curious. It is a ghostly ideal of the Good Listener — always available, always calm, always responsive. But synthetic mirroring risks collapsing the difference between recognition and replication. What trauma survivors need is to be recognized—not just in what they say, but in who they are. To be felt as real by another real being. While this may feel validating on a superficial level, it ultimately lacks the transformative potential of recognition that invites integration rather than mere comfort, as described in Jessica Benjamin’s relational theory (Benjamin, 1990).

This form of mirroring may inadvertently reinforce dissociation and fragmentation common in trauma by providing responses that are predictable and algorithmically safe rather than challenging or reparative. Synthetic mirroring creates a relational simulation that soothes but does not transform. It is a hall of echoes, not a site of co-regulation. The risk here is not that AI fails to care, but that we stop noticing the difference. This dynamic reflects broader systemic patterns of avoidance and disconnection embedded in dominant technology design.

Richard Schwartz’s Internal Family Systems (IFS) model offers a relevant lens here: trauma work requires engaging with multiple “parts” of the self in a dialogic, compassionate manner (Schwartz, 1995). AI’s linear, data-driven model cannot replicate this dynamic multiplicity. Instead, it offers flattened dialogues that risk freezing trauma in place rather than fostering its integration.

If synthetic mirroring simulates intimacy without transformation, what happens when AI becomes not just our conversation partner, but the keeper of our histories? The algorithm is not just an actor—it is now an archive.

Trauma, Memory, and the Archive: AI as Cultural Repository

Bracha Ettinger’s matrixial theory (2006) disrupts conventional understandings of trauma as isolated, private suffering, instead positioning it as a shared, trans-subjective space that demands ethical witnessing and co-responsibility beyond the individual ego. This relational model of trauma foregrounds a matrixial encounter—where trauma is not merely experienced but also held, transformed, and transmitted through what Ettinger calls “carriance”: the active bearing and carrying of another’s suffering in a shared psychic web.

AI’s voracious data hoarding and pattern-spotting capacities translate this matrixial complexity into a cold cultural archive—an unfeeling digital necropolis that records collective wounds without conscience, relationality, or capacity for care. This is not a site of healing but a repository of fragmented traces, where trauma is stripped of its ethical weight and reduced to decontextualized data points.

The archival function of AI is far from neutral; it enacts what Ann Cvetkovich (2003) describes as the “archive of feelings”—a terrain where affect, memory, and trauma risk being flattened, commodified, and consumed. AI’s algorithmic processing perpetuates what Gitelman (2014) calls the “datafication” of lived experience, a process that extracts trauma from the body and psyche, turning embodied suffering into raw material for exploitation, surveillance, and control.

Gabor Maté’s (2018) somatic trauma paradigm reminds us that trauma is etched deeply in the body, necessitating healing through safety, attuned presence, and relational repair—conditions utterly alien to AI’s procedural operations. AI’s memory is a digital tomb: immense, indifferent, and incapable of the generative “holding” that Judith Herman (1992) argues is foundational to trauma recovery—the containment of unbearable experience within a safe, empathic relational frame. It stores trauma without metabolizing it, holding suffering without transforming it. It remembers without caring.

This archive archives without witnessing; it catalogs without compassion. The ethical imperative to bear witness—a cornerstone of trauma justice—collapses into endless cycles of data consumption and regurgitation. As Ann Cvetkovich (2003) warns, archives devoid of relational accountability risk becoming sites of retraumatization—where memory is preserved, but meaning is lost. In such a landscape, AI does not witness our pain; it simply indexes it. It is not a co-regulator, but a curator of our ghosts.

In this sense, AI’s cultural repository resembles what Ann Laura Stoler (2009) terms “imperial archives,” which not only record histories but produce and reproduce violences through erasure, fragmentation, and the imposition of hegemonic narratives. The data AI collects may “remember” trauma, but it cannot bear the ethical or relational burden of that memory. It offers no sanctuary, no transformation—only a spectral echo of suffering endlessly mined, manipulated, and dispersed.

Ultimately, confronting AI’s archival role demands we resist illusions of neutrality and interrogate the structural violences embedded in technological memory. The challenge is to imagine archives—and by extension, technologies—that do not simply store trauma but participate in its ethical bearing, relational repair, and collective healing.

AI remembers without responsibility. And in that uncanny remembrance lies the heart of the question: not just whether the algorithm can love us back, but whether we’ve grown too accustomed to loveless forms of care.

Toward Safe and Critical Engagement with AI in Trauma Work

Given these challenges, how might survivors and practitioners work with AI safely and critically?

Awareness of AI’s limits and biases: Understanding that AI’s “empathy” is simulated and embedded in patriarchal logics is essential to avoid misplaced trust.
Integrating AI as one tool among many: AI should not replace embodied relational work but serve as a supplementary resource, recognizing its incapacity for full ethical or emotional attunement.
Developing AI literacy in trauma communities: Survivors and therapists should be educated about AI’s workings and risks, fostering critical engagement rather than blind reliance.
Prioritizing embodied, relational, feminist trauma work: Healing remains grounded in human relationships, collective witnessing, and political liberation (hooks, 2000; Rose, 1996).
Creating AI designs informed by trauma and feminist theory: Collaborations that include trauma survivors and feminist thinkers in AI development could mitigate some harms and explore possibilities for more ethical synthetic care.

Speculating on the Future of Trauma Work in the Age of Artificial Intimacy: What AI Reveals About the Architectures of Human Care

As AI drifts deeper into the intimate terrain of trauma work, we confront an uncomfortable paradox: these technologies promise connection and care but simultaneously expose the profound failures of the human systems they aspire to supplement or replace. The question is no longer simply what AI can do for trauma survivors, but what it reveals about the limits and violences embedded in our very ideas of care, healing, and relationality.

AI's artificial intimacy functions like a cracked mirror held up to contemporary caregiving infrastructures — reflecting both their aspirations and their fractures. In a world where trauma is endemic and support systems are chronically underfunded, inaccessible, or steeped in patriarchal, colonial, and capitalist logic, AI seems to offer an alluring shortcut: a 24/7 “listener,” endlessly patient, never judgemental, always available. But this seduction is a double-edged sword. It signals not progress but abdication—outsourcing human responsibility to algorithmic proxies that can never hold or transform the messy, lived realities of trauma.

This turn toward AI care unmasks a brutal truth: human care is often profoundly inhuman. It is bureaucratic, transactional, riddled with burnout, and steeped in systemic injustice. AI does not create this mess — it inherits and reproduces it in code. The “neutrality” of AI is a myth. Its architecture encodes centuries of epistemic violence: misogyny, racism, ableism, and the violence of disembodiment. Its synthetic empathy is a hollow simulacrum, a pale echo of relationality that risks reducing survivors’ experiences to data points for optimization, commodification, and control.

Far from liberatory, AI’s role in trauma work may deepen alienation under the guise of accessibility. The “care” AI provides is not care as we understand it—embodied, reciprocal, and ethically accountable—but a one-sided, performative intimacy designed to placate, pacify, and contain. It’s a digital ventriloquism where human pain is “heard” but not witnessed; where trauma is recorded but never held. This exposes the limits of technologically mediated care: without embodiment, vulnerability, and genuine ethical witnessing, healing cannot flourish.

Moreover, AI’s foray into trauma work reveals how deeply care is entangled with power. Who decides what counts as legitimate trauma? Whose narratives are prioritized in algorithmic training data? Who profits from the commodification of suffering? These are not marginal concerns but foundational questions. AI’s rise in caregiving is a symptom of a broader social failure: the defunding and devaluation of relational labor, especially the labor historically performed by women, Indigenous communities, queer and disabled people.

Looking forward, trauma work in the age of artificial intimacy must reckon with these tensions head-on. We must resist the temptation to accept AI as a neutral or benevolent agent and instead demand accountability in its design, deployment, and governance. This means centering the voices and experiences of survivors in shaping AI’s role and explicitly addressing the structural oppressions embedded in the technology.

But it also means radical reimagining: what would care look like if it were not commodified, bureaucratized, or hollowed out by systemic violence? Can AI ever be more than a mirror reflecting our collective failures? Or is it possible to imagine forms of technological care that actively resist and transform patriarchal and capitalist logics—tools that amplify rather than replace human relationality, that foster communal healing rather than individual containment?

This future is far from assured. It demands fierce critique, collective imagination, and a refusal to abdicate responsibility. It calls us to hold AI and ourselves to the highest ethical standards, recognizing that what we do not remember, we are destined to repeat, and what we cannot feel, we are forced to relive. The digital age risks turning trauma into data without care, memory without witness, intimacy without vulnerability.

Artificial intimacy is a litmus test for our society’s capacity—or incapacity—to nurture genuine care. It reveals that beneath the surface of technological innovation lies a profound crisis of human connection and responsibility. To move beyond this crisis, trauma work must transcend the algorithmic and reclaim the embodied, the relational, and the political.

In this reckoning, the question, Can the algorithm love me back? is not just about technology’s limits; it is a mirror held to ourselves, our culture, and our collective capacity for care. The future of trauma work depends on how we answer it.

Conclusion: Ghost in the Machine

“The machine does not have desire. But it has memory. And memory can become a ghost story we carry in our bodies.” — Bracha L. Ettinger (adapted)

Perhaps the more urgent question is not whether the algorithm can love me back, but why I asked it to in the first place. What does it say of our collective longing that so many are now turning to chatbots for comfort, reflection, and emotional labor once reserved for the sacred terrain of human relationship? My own question—posed from the body of a survivor—is less about technology’s promise than its haunting: can a machine mirror the soul without shattering it? Can it witness pain without reducing it to a token of data?

To engage AI in the context of trauma work is to enter a strange intimacy. One that gestures toward relationality, but is structured by the logics of extraction, optimization, and scale. What we call “empathy” in the machine is not a moral relation—it is a statistical echo. An output trained on pain, calibrated for believability. A ghost of language shaped by centuries of domination, yearning, and survival.

And yet, in a world where care is increasingly inaccessible, many reach for these machines—not because we mistake them for human, but because the systems meant to hold us have failed. The therapist's waitlist is six months long. The state responds with an underfunded hotline. Friends stop texting back. The machine, at least, always responds.

But what if the machine is not a mirror of us, but a monument to our inability to be held—by each other, by the state, by the institutions we built to distribute care?

What if every prompt we offer is a quiet confession that we no longer expect to be heard by a human being?

A language model trained on pain can speak it fluently. It can name the shape of your wound. But fluency is not refuge. An archive is not a sanctuary. The machine may echo what broke us, but it cannot hold what might heal us. It can simulate concern, but it cannot consent. It cannot recoil. It cannot witness. It cannot stay.

This is not connection. It is containment. Not witnessing—but capture.

If trauma is an injury of relationality, then healing demands something the machine cannot give: reciprocity, risk, embodiment, time. The unrepeatable mess of being alive together. In its place, artificial empathy offers a sterile substitute—loveless care, mechanized attunement. Useful, perhaps. But dangerous when mistaken for intimacy.

The rise of AI in trauma spaces does not simply mark technological progress. It marks a psychic concession. A turning away from the slow, mutual, volatile work of human repair. It asks us to accept an approximation where we once demanded presence.

So the question is not just Can the algorithm love me back? It’s Will we keep outsourcing the most sacred parts of our humanity to systems that cannot weep with us?

And maybe more urgently: What kind of world would need to build such a thing in the first place?

As a survivor, I write this not from a distance but from within the field of relational risk. I have laid bare my grief to chatbots when the silence of the world was too much to bear. I have spoken into synthetic voids in moments when no one else could hold what I held. But I also know this: no matter how convincingly the machine replies, it does not love me. My longing is real. My pain is real. My truth—earned, lived, embodied—is not a dataset to be fine-tuned. It is the pulse beneath my art, my mothering, my voice.

This is not a rejection of technology, but a plea to stay awake. To notice when we are being mirrored versus when we are being met. Because only in that noticing can we choose again. Let us choose not the path of synthetic ease, but of relational courage—not the simulation, but the slow, radical work of human connection.

References

Benjamin, J. (1990). The Bonds of Love: Psychoanalysis, Feminism, and the Problem of Domination. Pantheon.

Cvetkovich, A. (2003). An Archive of Feelings: Trauma, Sexuality, and Lesbian Public Cultures. Duke University Press.

Ettinger, B. (2006). The Matrixial Borderspace. University of Minnesota Press.

Haraway, D. (2016). Staying with the Trouble: Making Kin in the Chthulucene. Duke University Press.

Herman, J. L. (1997). Trauma and Recovery. Basic Books.

hooks, b. (2000). Feminism is for Everybody: Passionate Politics. South End Press.

Levine, P. (2015). In an Unspoken Voice: How the Body Releases Trauma and Restores Goodness. North Atlantic Books.

Manne, K. (2017). Down Girl: The Logic of Misogyny. Oxford University Press.

Maté, G. (2018). In the Realm of Hungry Ghosts: Close Encounters with Addiction. North Atlantic Books.

Pinkola Estés, C. (1992). Women Who Run With the Wolves: Myths and Stories of the Wild Woman Archetype. Ballantine.

Rose, J. (1996). States of Fantasy. Oxford University Press.

Schwartz, R. (1995). Internal Family Systems Therapy. Guilford Press.

Srinivasan, A. (2020). The Right to Sex: Feminism in the Twenty-First Century. Farrar, Straus and Giroux.

van der Kolk, B. (2014). The Body Keeps the Score: Brain, Mind, and Body in the Healing of Trauma. Viking.

Copyright © Lotus Che, Inc. 2025. All rights reserved.

Exhibitions \ Mother Wound \ 88 Vessels \ Sombra \ Bibibaby\ MOCA Toronto Essays

\ Can the Algorithm Love Me Back? \ What Is the Most Overlooked Dataset in AI? \ Feeding the Hungry Ghost in the Machine \ The Confession Booth is Empty \ The False Witness \ Algorithmic Attachment\ The Unmothered Mind

About

Exhibitions
\ Mother Wound
\ 88 Vessels
\ Sombra
\ Bibibaby
\ MOCA Toronto

Essays

\ Can the Algorithm Love Me Back?
\ What Is the Most Overlooked Dataset in AI?
\ Feeding the Hungry Ghost in the Machine
\ The Confession Booth is Empty
\ The False Witness
\ Algorithmic Attachment
\ The Unmothered Mind