Equinox Withdraws AI Ad After Asian Creators Call Out Racist, Sexualized Imagery

Table of Contents

1. Key Highlights
2. Introduction
3. What the ad showed and how creators responded
4. The campaign concept and where it failed
5. Why that image hurt: historical stereotypes and modern consequences
6. Agencies, clients, and accountability: who approved what?
7. The role of AI in advertising: technical risks and blind spots
8. Real-world comparisons: other brand missteps and what they show
9. When apologies fall short: performative gestures versus structural change
10. Practical steps for advertisers and agencies
11. The power of creator communities and social accountability
12. Industry-level solutions and regulatory context
13. Short- and long-term consequences for brands
14. How to move beyond optics to meaningful change
15. Final reflections on accountability and creativity
16. FAQ

Key Highlights

• Equinox pulled an AI-generated out-of-home ad and issued an apology after Asian creators and advocates condemned the image as dehumanizing and fetishistic; the work was produced with agency Angry Gods.
• The campaign intended to critique AI fakery but included an eroticized, AI-generated depiction of an Asian woman that invoked long-standing racialized stereotypes; the agency pledged donations and edited its apology after online backlash.
• The episode highlights structural failures—insular creative review, inadequate cultural competency, and unexamined reliance on AI—offering concrete lessons for marketers, agencies, and ad platforms.

Introduction

A luxury fitness brand launched a campaign that aimed to question the authority of digitally manufactured images. Instead of drawing attention to the perils of fabricated content, one execution amplified a historical pattern of racialized sexualization. The ad, which juxtaposed surreal AI creations with photographs of real people, featured an artificial image of an Asian woman’s head mounted on a rod, with a finger pressing a thumb between her lips. Creators on Instagram and TikTok called the image dehumanizing, a literal embodiment of stereotypes that have positioned Asian women as objects, dolls and hypersexualized fantasies for generations.

Equinox removed the poster from displays in New York, Los Angeles and San Francisco after sustained criticism and apologized. Its creative partner, Angry Gods, acknowledged the harm, edited an initial statement after receiving pushback, and pledged a charitable donation. The incident exposes how the fast adoption of AI tools and a lack of diverse scrutiny can produce work that, regardless of intent, perpetuates real-world harms. It also raises questions about accountability in advertising: how brands approve content, how agencies think about cultural context, and what policies are necessary to prevent similar lapses.

This article reconstructs what happened, analyzes why the image struck a nerve, places the controversy in a wider pattern of brand missteps, and lays out practical reforms for advertisers who want to use emerging technologies without repeating old harms.

What the ad showed and how creators responded

The contested image was part of Equinox’s “Question Everything But Yourself” campaign, which rolled out in January. The broader concept used absurd, AI-generated images—like Pope Francis in a puffer jacket and an exaggeratedly posed Justin Trudeau—to draw a contrast between digitally fabricated fantasies and human reality. One poster from the campaign paired an AI-rendered image of an Asian woman’s head, isolated and mounted on a metal rod, with a photograph of a different woman whose torso was partially exposed.

Creators reacted quickly and forcefully. Artist and content creator Eunnuri Lee framed the image in historical context: "Asian women have already been treated as dolls, submissive sex machines, anime girls, robots, decorative objects [and] mail-order fantasies for decades now," she said in a video. “We are constantly positioned as hypersexual beings and also not fully human.” Food creator Soogia, who has nearly three million followers, called the ad a “literal embodiment” of the sexualization and objectification of Asian women. Other creators, including @relishwithreese, described the image as “disturbing.”

Ed Choi, a creator and podcaster, directly criticized Equinox’s chief marketing and digital officer in a video, identifying the image as an example of oversexualizing Asian women. These responses converged on a single point: the ad did not read as surreal or playful in the same fashion as the rest of the campaign. Instead, it trafficked in centuries-old depictions that strip away humanity and agency.

Equinox initially responded by stating it was “in the process” of removing the image from out-of-home placements. The company confirmed the ad was taken down from 10 club locations in New York, Los Angeles and San Francisco where it was still visible. An Equinox spokesperson said the campaign aimed to critique “false realities created by AI and social media,” but acknowledged “an unintended impact” and apologized, citing core values of inclusivity, mutual respect and empathy.

Angry Gods, the Los Angeles-based design and strategy agency behind the work, issued an apology on Instagram. After receiving criticism for downplaying the backlash as a “discussion,” the agency edited its statement and committed a donation of $5,000 to the National Asian Pacific American Women’s Forum; the agency’s founder and CEO added $5,000 personally.

The campaign concept and where it failed

The campaign’s creative brief centered on an idea that resonated with marketers: AI is generating convincing but hollow imagery that can distort reality. By pairing intentionally absurd AI images with photographs of real people, the campaign sought to remind audiences of the difference between manufactured spectacle and authentic human strength. That conceptual clarity held promise. The execution failed because one of the surreal images borrowed, inadvertently or not, from a visual vocabulary steeped in racialized sexual stereotypes.

The team’s failure occurred at several levels:

• Prompting and model outputs: If the AI model was prompted with stereotyped descriptors—or trained on images that replicate fetishistic portrayals—produced outputs would reflect those biases. Creatives who accept the output without probing how it was generated or considering its socio-cultural weight risk amplifying harm.
• Inadequate review pipelines: The campaign made it to large-format out-of-home placements and social channels, indicating the work passed multiple sign-offs. That suggests review processes did not include diverse perspectives or external checks for cultural harm.
• Misalignment of tone: The campaign mixed playful absurdity with a rendering that read as eroticized and violent. The incongruity between the image and the campaign’s stated aim made the harm more visible. Where other images in the campaign skewed toward satire, the objectifying depiction stood out—and not in a way that advanced the campaign’s critique.
• Blind spots in cultural competency: Angry Gods later acknowledged that “an artificial, eroticized Asian woman’s face is not a neutral symbol of ‘fake’,” and that the agency “should have seen that before the work ever shipped.” That admission points to an absence of historical awareness and a failure to involve voices who could have flagged the image before launch.

The irony is acute. A campaign designed to call out “fake” imagery relied on a caricature that perpetuated a real, systemic pattern of erasure and objectification.

Why that image hurt: historical stereotypes and modern consequences

The image resonated for thousands because it did not exist in a vacuum. Visual stereotypes of Asian women have long histories and manifest in overlapping tropes: the submissive “lotus blossom,” the hypersexual “sex worker” or “mail-order bride,” and the fetishized figures of anime and dollification. These archetypes appear across literature, film and advertising, shaping how broader audiences perceive Asian women’s agency, desirability and humanity.

Stereotypes translate into lived consequences. Sexualized tropes contribute to harassment and disproportionate targeting in violent crimes. Studies and surveys of Asian American experiences document elevated rates of sexual harassment and a particular vulnerability of Asian women to racially motivated sexual violence. Stereotypes also constrain career opportunities and social mobility by funneling perceptions of worth into narrow, sexualized frames.

The visual in the Equinox ad layered those stereotypes—an artificial, objectified head with a suggestive gesture—into public space. The difference between a provocative art piece and an unexamined stereotype lies in context and consent. Art that deploys charged imagery tends to do so with a clear framing, intent, and conversation about its provocation. Advertising machines that use similar visual shorthand do not afford the public the same interpretive space; ads leverage familiarity and routinized exposure to normalize images at scale.

Creators who pushed back were not calling for censorship of experimentation. They demanded accountability: a recognition that certain tropes carry harm and an understanding of how those tropes function within a history of racialized gendering. The image triggered a cascade of painful cultural memory. It read, to many viewers, as another iteration of a pattern that reduces Asian women to objects.

Agencies, clients, and accountability: who approved what?

Advertising production involves multiple layers: a brief from a client, creative ideation by an agency, design and production, legal and brand review, media buys, and out-of-home vendors preparing placements. At any of those junctures, a problematic image can be caught—if the right people and processes are in place.

In this case, the agency took responsibility for the mistake and pledged donations, but the timeline suggests the ad passed several approval points, including Equinox’s internal brand team and external media placements that allowed the poster to remain up for months in some cities before removal. Equinox deleted a video showing the ad prominently displayed in a New York window after criticism intensified.

Accountability in advertising must be clear: clients cannot outsource responsibility for cultural competence. Agencies do not get sole latitude to experiment without client sign-off. Brands that position themselves as inclusive or progressive have an additional obligation to ensure marketing decisions align with their public values. Equinox framed its response around core values but did not provide granular detail on how the image was approved, who saw it, and what steps will change to prevent recurrence.

Angry Gods initially characterized the backlash as a “discussion,” a framing that minimized harm and underestimated the urgency of the response. The agency later edited that language and acknowledged the historical freight an eroticized Asian face carries. That shift reflects a learning moment, but critics often view apologies filtered through PR cycles as incremental and reactive unless followed by sustained structural change.

The donation to the National Asian Pacific American Women’s Forum is a tangible step. Still, many advocates were clear that charitable giving must coexist with substantive reforms—hiring practices, meaningful partnerships with Asian American organizations, and transparent audits of creative processes.

The role of AI in advertising: technical risks and blind spots

AI tools are accelerating creative production. They can generate prompts, imagery and variations at scale, enabling faster ideation. Those capabilities drive interest among brands and agencies eager to appear innovative. However, the very mechanics that make AI useful also create risks.

AI models learn from large datasets scraped from the internet. If those datasets contain biased, fetishized or otherwise harmful content, the models absorb and reproduce those patterns. Specific risks include:

• Bias amplification: AI systems can entrench and amplify existing biases in training data. Racialized or sexualized depictions present in source material will be reflected and potentially exaggerated in outputs.
• Hallucination and lack of provenance: Creative teams may not know the exact provenance of generated images. Without accountability for how AI arrived at a result, harmful associations can slip through.
• Lack of nuance in prompts: Using shorthand, slang, or ambiguous modifiers in prompts may produce images that carry unintended meanings. Even neutral prompts can yield problematic outputs if the model equates certain descriptors with stereotypical imagery.
• Overreliance on automation: Accepting AI outputs at face value without human-in-the-loop review is dangerous. Human review must include cultural competency as well as aesthetic critique.

To reduce these risks, teams must adapt workflows. Agencies and brands should implement prompt audits: document the inputs used to generate imagery and subject them to review. Maintain a human-in-loop model where diverse stakeholders inspect and sign off on AI outputs. Use AI tools that provide transparency about training data and content filters. Establish red lines that define prohibited imagery and scenarios, including fetishization of protected groups, dehumanizing depictions, or uses that echo historical trauma.

Technology companies that provide generative models also bear responsibility. Clearer documentation on training datasets, content filters tuned to cultural harms, and explainability features that help creatives understand why a model produced a given output would reduce downstream risk. Advertising agencies should require vendors to disclose such features before onboarding.

Real-world comparisons: other brand missteps and what they show

The Equinox episode is not an isolated incident. Brands have repeatedly misjudged imagery and messaging, sometimes with immediate commercial consequences, sometimes with reputational impacts that linger.

American Eagle’s campaign featuring Sydney Sweeney—referenced in coverage of the Equinox incident—drew criticism for imagery that some described as racist and overly sexual. American Eagle publicly stood by the campaign, and internal comments later suggested the controversy produced a business uplift. That example highlights a painful reality: controversy can generate attention that translates into short-term sales even as it damages trust with specific communities.

The fashion and luxury industries have a long history of cultural missteps. In another well-known case, a major brand’s advertising in China was criticized for belittling local customs and norms; the fallout included canceled events and public apologies. These incidents share a common thread: brands with global reach must navigate cultural differences and histories. Creative teams that treat cultural markers as merely aesthetic risk repeating harm.

Each incident demonstrates that outcomes vary. Some brands weather backlash and see immediate commercial gains. Others face long-term erosion of brand equity and consumer trust. The determining factor is rarely a single ad; it is how the brand listens, responds, and adapts its internal structures to prevent repetition.

When apologies fall short: performative gestures versus structural change

Apologies are a necessary first step. They acknowledge harm and create a public record that the brand recognizes the problem. They are not, by themselves, sufficient.

A meaningful response requires:

• Transparency about what went wrong: Brands should explain the chain of approvals, including whether AI tools were used, who signed off, and how the image was allowed to publish.
• Concrete remedial steps: Commitments to specific changes—revamping review processes, hiring diverse creative leads, convening external advisory panels—signal a shift beyond PR.
• Long-term investment: Donations to relevant advocacy groups are positive. They must be accompanied by ongoing partnerships and financial commitments that extend beyond a single, reactive payment.
• Metrics and accountability: Public reporting on progress—such as diversity metrics within creative teams, audits of AI outputs, and the number of projects reviewed by cultural consultants—creates accountability.

The Equinox and Angry Gods statements included removal of the ad, an apology, and a small donation. That response addresses immediate exposure but leaves questions open about process reform. Many viewers and industry watchers will look for follow-up: Did Equinox change its approval workflow? Will Angry Gods institute independent cultural review? How will future campaigns be audited?

Brands that treat accountability as a one-time exercise will likely face renewed criticism when the same gaps reappear. Firms that embed cultural review into production pipelines and publish ongoing progress reports are more likely to rebuild trust.

Practical steps for advertisers and agencies

Advertising teams that want to use AI responsibly must translate high-level commitments into everyday practices. The following measures reduce risk and improve creative outcomes:

1. Establish clear AI-use policies: Define what tools are authorized, which use cases are prohibited, and how prompts and outputs must be documented. Include clauses that prohibit fetishizing or dehumanizing portrayals of protected groups.
2. Document prompts and outputs: Maintain an auditable log of prompts and resulting images. This traceability supports internal review and helps explain decisions if issues arise.
3. Create a human-in-the-loop review process: Ensure a diverse panel reviews AI-generated assets before approval. Include subject-matter experts on race, gender and cultural history where content touches those areas.
4. Use external advisory boards: Engage community organizations, cultural critics, and advocacy groups for pre-launch feedback on campaigns that intersect with identity and representation.
5. Train creative teams in cultural literacy: Invest in continuous education about stereotypes, historical tropes, and contemporary harms. Training should be mandatory for those involved in creative production and sign-off.
6. Test with representative audiences: Use focus groups or small-scale test placements in diverse communities to detect unintended readings before wide release.
7. Maintain vendor standards: Require agencies and technology vendors to have their own content moderation and transparency practices and to disclose training data provenance when possible.
8. Define red lines and style guides: Create brand guidelines that explicitly forbid certain imagery (e.g., depictions that dehumanize, sexualize minors, or use racialized caricature). Add examples to clarify gray areas.
9. Invest in long-term partnerships: Instead of one-off donations, cultivate sustained relationships with advocacy organizations and creators from underrepresented groups. Collaboration builds trust and provides real-cultural insight.
10. Publish accountability metrics: Share progress publicly—diversity of creative teams, number of campaigns reviewed by cultural advisors, incidents reported and remediated.

These steps are not exhaustive, but they convert abstract principles into operational guardrails.

The power of creator communities and social accountability

Digital creators played a decisive role in this story. TikTok and Instagram creators amplified the objection, framed the issue in historical context, and pressured both agency and client to act. The dynamic demonstrates the shifting power balance in media: influencers and creators are not just distribution channels; they act as cultural interpreters and watchdogs.

Creator-led accountability tends to be fast and visible. Videos that distill complex historical critiques into concise, emotionally resonant narratives often reach far beyond the communities directly affected. Brands must recognize creators as stakeholders, not just promoters. When campaigns touch on identity, involving creators early—especially those from impacted communities—can surface problems and enrich the creative strategy.

Creators also serve an evidentiary function: their commentary documents community harm and helps the public understand why a particular image causes offense. That clarity matters in a crowded media environment where single images can be viewed without context. Creator-led critique compels brands to respond, and sustained creator pressure is often the difference between a perfunctory apology and substantive reform.

Industry-level solutions and regulatory context

The advertising industry lacks a universal standard for the ethical use of AI-generated content. Trade associations, ad networks and regulators can help close that gap.

Industry groups can develop best-practice guidelines that cover AI use in advertising, including requirements for transparency, provenance, and cultural review. Shared standards would prevent a patchwork of policies that leave smaller agencies vulnerable to mistakes.

Regulators are increasingly attentive. Emerging frameworks—such as the European Union’s AI Act and national-level proposals—aim to increase oversight of high-risk AI applications. Advertising that uses AI to create realistic people or manipulate identity may intersect with these regulatory trajectories, especially when harmful content targets protected groups.

Out-of-home vendors and platform operators also have power. Vendors can refuse placements that violate community standards, and platforms can require documentation for AI-generated content. These gatekeeping functions create new accountability touchpoints in the media supply chain.

Self-regulation remains central. Agencies and brands that seize the initiative to publish responsible use policies and to submit to independent audits will raise the bar across the industry.

Short- and long-term consequences for brands

A single misstep can have varied consequences. Short-term effects often include social media blowback, calls for boycott, or cancellation of sponsored content. Long-term effects are harder to quantify but include erosion of trust among targeted communities and potential impacts on recruitment and partnerships.

Some brands emerge from controversy with increased visibility and a temporary spike in attention, even sales. That pattern illustrates a perverse incentive: controversy can deliver immediate metrics that appease stakeholders focused on growth. The tradeoff is reputational capital among communities whose support is essential for sustained brand health.

Companies that make a series of misjudgments may experience compounding harm. Reputation is cumulative. A brand that consistently fails to demonstrate cultural competence will find it harder to hire diverse talent, earn media trust, or partner with organizations that advance inclusion. The cost of repair rises with each failure.

Equinox’s brand positioning—as a premium fitness club with a cultural cachet—makes credibility on values like inclusivity particularly important. The way the company addresses the mistake will influence whether the damage is shallow and recoverable or a deeper wound to brand identity.

How to move beyond optics to meaningful change

Performative measures—apology posts and one-off donations—are inadequate if organizations fail to embed new practices. Meaningful change requires structural investments:

• Integrate cultural review into procurement: Require agencies to demonstrate how they will prevent cultural harm as part of contracting.
• Budget for ethical oversight: Build fees for external cultural audits and inclusive testing into campaign budgets.
• Elevate diverse decision-makers: Position staff from underrepresented backgrounds in roles with sign-off authority on creative direction.
• Track and publish progress: Report on the implementation of new review processes and their outcomes.
• Partner with community organizations: Launch collaborative campaigns with advocacy groups, providing resources and creative voice rather than using communities only as validators after a mistake.

Brands that adopt these measures will not merely react; they will create resilient workflows that reduce the likelihood of repeating harm while enabling responsible experimentation with technology.

Final reflections on accountability and creativity

Creativity thrives on risk and provocation, but that freedom carries moral obligations when deployable at scale. Advertising lives in public space; images on billboards, in subways and on social feeds shape cultural narratives. AI will continue to expand what creatives can produce. The crucial test for brands and agencies will not be whether they use AI, but how they use it.

The Equinox episode serves as a case study: a strong concept, a provocative device and a failure of cultural attunement. The apology and removal addressed immediate harms, but the deeper work is systemic. Agencies must bring cultural literacy to technical experimentation. Clients must demand transparency and set clear boundaries. Creators and advocacy groups must be treated as stakeholders with expertise.

When brands build processes that treat representation as integral to creativity—rather than an afterthought—the industry will produce work that is both inventive and responsible. That alignment preserves creative ambition while guarding against the repetition of harm that communities experience far too often.

FAQ

Q: What specifically made the Equinox ad offensive? A: The ad used an AI-generated image of an Asian woman’s head presented in an objectifying, dehumanized manner—mounted on a rod and shown with a thumb between her lips. That composition evoked historical tropes that fetishize and sexualize Asian women. Advocacy and creator responses emphasized that the image reinforced patterns of racialized gendering rather than functioning as neutral satire.

Q: Did Equinox or Angry Gods apologize and take responsibility? A: Yes. Equinox removed the image from out-of-home placements in several cities and issued an apology that acknowledged “unintended impact.” Angry Gods apologized, edited an initial statement after criticism, and pledged a $10,000 donation to the National Asian Pacific American Women’s Forum ($5,000 from the agency and $5,000 from the agency’s founder).

Q: How did the image make it into public placements? A: The campaign was produced with standard agency-client workflows and deployed across social, digital and out-of-home placements. The presence of the ad in public spaces for months indicates it passed multiple approvals. The precise chain of sign-offs has not been fully published, which is a key question for accountability.

Q: Is using AI in creative work inherently risky? A: AI can offer powerful creative tools, but risks arise when models reproduce or amplify biased content from their training sets. The technology is not inherently malicious, but it requires careful human oversight, transparent workflows and cultural review processes to avoid replicating harmful stereotypes.

Q: What should agencies and brands do to prevent similar incidents? A: Implement clear AI-use policies; document prompts and outputs; create human-in-the-loop review processes that include diverse perspectives; use external cultural advisors; require vendor transparency about model training where possible; conduct representative audience testing; and invest in cultural competency training. Publicly commit to measurable changes and follow through with reporting.

Q: Are donations sufficient as remediation? A: Donations are one component of remediation but are not sufficient on their own. Meaningful remediation includes transparent process changes, ongoing collaboration with community groups, structural reforms in hiring and creative approval, and measurable accountability that reduces the likelihood of repeat offenses.

Q: Could regulatory frameworks affect future use of AI in advertising? A: Yes. Emerging AI regulations, such as provisions in frameworks like the EU AI Act and other national policies, aim to govern how AI is deployed, particularly in high-risk contexts. Advertising that uses AI to depict people or manipulate identities may increasingly fall under regulatory scrutiny.

Q: How can consumers hold brands accountable beyond social media criticism? A: Consumers can seek transparency by asking brands directly about their review processes, sign-on to petitions that call for systemic changes, support advocacy groups that push for industry standards, and evaluate brands’ follow-through on commitments by tracking public reporting and long-term behavior.

Q: Does controversy always hurt a brand’s business? A: Not always in the short term. Some brands see temporary spikes in attention and sales after controversy. Long-term effects are mixed and depend on the brand’s response, consistency of behavior, and whether the brand earns back trust with affected communities. Repeated missteps can erode brand equity permanently.

Q: What role should creators and community voices play in future campaigns? A: Creators and community stakeholders should be engaged early and meaningfully in campaigns that touch on identity and representation. Their participation should be compensated and their feedback integrated into creative development, not treated as a post-launch validation mechanism.