Upon seeing my colleague Kylie Robison’s article about OpenAI’s latest image generator earlier this week, I anticipated an enjoyable few days. Though generative AI images prompt numerous ethical debates, their creativity captivated me, and I spent a fair amount of time observing fellow Verge staff push ChatGPT’s capabilities in uniquely entertaining ways.
However, Thursday took an unexpected turn when the White House posted a provocative photograph of a crying detainee, claiming she was an arrested fentanyl trafficker and undocumented immigrant. Accompanying this image was what appeared to be an AI-generated cartoon depicting a law enforcement officer handcuffing the distraught woman. The illustration bore a strong resemblance to ChatGPT’s popular imitations of Studio Ghibli animation, which have recently flooded social media.
While the inappropriate use of the tool shouldn’t tarnish its reputation entirely, the combination of the charming Ghibli aesthetic with the White House’s marketing effort felt disturbingly appropriate. Both seem to emerge from a vein of thought that dismisses basic human decency as a sign of weakness, favoring a cold, authoritarian disposition instead.
We have reached out to OpenAI and the White House for clarification, but this maneuver appeared to be a strange promotional tactic for a company closely associated with President Donald Trump. Political leaders have long leaned into internet memes, and while it remains uncertain whether ChatGPT or another AI model produced the image, the possibility cannot be ignored. OpenAI’s CEO, Sam Altman, has previously showcased the Ghibli-style images as an exclusive feature for subscribers to ChatGPT’s premium service. Trump has frequently endorsed OpenAI’s Stargate initiative, even unveiling it at a press conference with Altman.
The association between the animated style and the former president may seem incongruent at first glance. The intent behind the White House’s post appeared to embody a form of cruel and performative governance, reminiscent of past controversial social media moments, like the “ASMR: Illegal Alien Deportation Flight” video featuring images of clinking chains. Even if the claims made in the content are accurate, it presents a troubling image of public humiliation directed at an immigration detainee, which few would regard as effective governance or moral integrity.
The wholesome nature of the Ghibli aesthetic feels jarring against this backdrop. Even prominent conservative voices from Silicon Valley have noted that visualizing a tearful anime woman being apprehended by an unyielding agent does little to curry favor with the public for such authority figures.
The use of AI-generated media aligns closely with elements of the MAGA movement, which has produced various unsettling and inappropriate creations. This trend stems from a deep-seated affinity for hyperbolized images of Trump in political satire as well as connections between the former president and the AI sector, notably through figures like Elon Musk and projects like Stargate.
Eight years ago, a tech firm might have distanced itself from individuals leveraging its memes to advocate for mass deportations.
It is unclear how OpenAI and Altman view this public endorsement of a brutal campaign against immigrants in connection with their promotional content. Historically, Altman supported progressive causes until the current administration took power. Before this controversial post, OpenAI had indicated that ChatGPT’s image generation features would have stringent guidelines, which might prompt them to argue this incident is akin to the misuse of Photoshop. If this image were not linked to the White House, some might interpret it as a critique of the arrests.
However, in a markedly different political climate from eight years ago, a prominent tech enterprise would likely distance itself from such sentiments. It wouldn’t be far-fetched for a company to assert that “OpenAI values artistic freedom and user responsiveness, but this administration’s stance does not align with our principles.”
The pressure to refrain from expressing such dissent today is formidable. Regardless of personal views within OpenAI, overt criticism of a vengeful administration seems unwise, especially amid a growing shift in Silicon Valley towards right-leaning ideologies.
Beneath this tension lies an ideological struggle, as the Ghibli filter embodies a stark reminder of the current era’s insensitivity to the humanity of others.
I personally appreciate the charm of Ghibli-tinged imagery. Many creations bring undeniable joy, and anime filters have gained popularity for their appeal. Yet, renowned filmmaker Hayao Miyazaki, synonymous with the Studio Ghibli brand, has outspokenly criticized AI art, famously denouncing earlier iterations of AI animation as “an insult to life itself.” It is unlikely he condones the imitation of his artistry through ChatGPT, nor does it seem appropriate for OpenAI to profit from it. Appropriating Ghibli’s distinctive style for marketing purposes, as noted by author Brian Merchant, sends a clear signal to the artists behind the AI’s development: We will take your work without seeking your consent.
OpenAI could have chosen to view artists as collaborators, not merely as sources of training data.
The intersection of technology and politics reveals a dominant ideology that prioritizes power—forcing compliance through strength and authority. This is evident in Trump’s approach, but it also relates to technology that favors superficial innovations over meaningful progress. The propensity to view criticism as hollow or adversarial while dismissing empathy as weakness speaks volumes about the current landscape.
Ironically, amidst a plethora of trivial AI applications, the Ghibli filter has emerged as a fan favorite. There exists a scenario where OpenAI could harness this popularity without undermining the artistic community that underpins its technology. By collaborating with artists rather than viewing them as expendable, AI developers could produce enriching partnerships that benefit all. Even if a prominent figure like Miyazaki might resist endorsement, there are many lesser-known animators whose work could be showcased instead, fostering a more inclusive artistic environment based on mutual respect.
Lastly, while purchasing a subscription to ChatGPT doesn’t inherently define someone’s morality, it prompts reflection on ethical consumption. Ultimately, we navigate our own boundaries regarding support for various practices in the tech industry.
The recent Tesla protests illustrate how aligning with controversial political narratives can have unintended negative consequences. Although many utilize ChatGPT for creating delightful portrayals, it is disheartening to witness the platform being exploited by the White House to undermine the marginalized. As OpenAI’s mission reflects a commitment to “AI for good,” how do its actions resonate with this vision? Moreover, in an increasingly competitive AI market, where will companies like OpenAI draw their ethical lines?
