Stanford admits using AI to race swap white male student for black female in ad: ‘Erased’

Read the original at New York Post ↗
Cite

Khan, Z. (2026, September 24). Stanford admits using AI to race swap white male student for black female in ad: ‘Erased’. New York Post. https://nypost.com/2026/09/24/us-news/stanford-admits-to-ai-race-and-gender-swap-in-doctored-student-ad/

Dated when The Aporia collected the article, usually within hours of publication. Check the original for its exact date and byline.

New York Post · collected 2026-09-24 · by Zain Khan

Quick Summary

Stanford University admitted using artificial intelligence to alter a promotional image, replacing a white male student with an AI-generated black female figure in violation of its own policies. The incident came to light when the Stanford Review published side-by-side comparisons showing the original photograph and the altered version. Stanford acknowledged that the manipulation erased the identity of one student, Billy Ramirez, and distorted others' appearances, violating university guidelines which prohibit such practices without disclosure. The university is now investigating the matter and plans to provide additional training for staff to prevent similar issues in the future.
Written locally by qwen2.5:14b on 2026-09-24, using this article's own text rather than the other coverage of the same event (that is the story summary below).

AI analysis runs on qwen2.5:14b, locally

Story summary

Stanford University admitted using artificial intelligence (AI) to alter a promotional image of students, transforming a Hispanic male student named Billy Ramirez into a black woman and slimming other students' appearances. The altered photo was published on "Welcome Home" banners around campus by Stanford Residential & Dining Enterprises without disclosing the AI manipulation, violating the university’s own policy against using AI to alter images of people associated with the institution. After The Stanford Review, the student-run newspaper, reported the incident, the university acknowledged it as a serious error in judgment and said they were working on additional training for staff involved. Ramirez expressed initial confusion followed by upset upon seeing his erasure through AI technology.

Written for “Stanford AI Race Swap Controversy” on 2026-10-16, grounded in this article and the 3 other(s) covering the same event.

Story

📰 Stanford AI Race Swap Controversy
Education · 4 article(s) covering the same event. Compare every outlet's account →

Compared with similar articles

This article reads unscored and hedges 8% of its claims. Each row says how that neighbour differs.
CBS News
⚖️ leaning not scored 🔴 0% hedged 0 of 3 📰 publisher trust 66
“Both articles report on Stanford University's admission and apology for using AI to manipulate a promotional image in violation of their policies, specifically altering the race and gender of a student.”
NBC News · 0.88 cosine similarity
⚖️ Leans left 🔴 10% hedged 2 of 21 📰 publisher trust 95
“Both articles describe Stanford University's admission and apology regarding the use of AI to alter a promotional image, changing a male student's appearance to that of a Black woman in violation of university policy.”
The Independent · 0.88 cosine similarity
⚖️ leaning not scored 🔴 12% hedged 3 of 25 📰 publisher trust 59
“The articles describe different races for the student replaced in the AI-altered image (Black woman replacing white male vs. Hispanic senior).”
How this article was measured signals, leaning charts, publisher record, byline, topics and claims

Signals How these are calculated →

Claims extracted
25
sentences that state something as happening or true
Uncertain
8%
2 of 25 hedged ("reportedly", "may", "according to")
Leaning
not political
takes no side on a contested political question
Publisher signals
67/100
New York Post's corrections and hedging, across everything collected from it. Not a measure of accuracy.
Outlets on this story
4
Education
Same framing elsewhere
1
articles that frame the event the way this one does
Analyzed 2026-09-24 · how these are computed

How this is being covered How these are calculated →

Article leaning vs. publisher reliability
Source leaning vs. consistency

Publisher

New York Post · 10960 article(s) · 31 correction(s) detected
Running correction rate · 31 correction(s)

Who wrote this

Zain Khan
51 article(s) here · 1 carrying a prediction
🔮 “The most experienced and the best teachers will often choose to work in more advantaged schools,” Polikoff said.
🔮 I don’t think no one in their right mind would jump off the fourth floor,” Amaya said.
🔮 Eight structures along the city’s Peninsula have been red-tagged as unsafe to enter, while another was yellow-tagged, allowing residents limited access to retrieve belongings but prohibiting occupancy until the property is deemed safe, Mayor Rex Richardson said.
🔮 When asked whether he supported a professor-led push to reinstate the SAT, Barrera said the test itself would not solve the underlying problem.
🔮 Los Angeles Unified School District is facing a staggering financial crisis that could force it to slash thousands of jobs over the next three years as California officials warn the nation’s second-largest school district is at “high” risk of insolvency.
🔮 I don’t think Israel will fall because of a bunch of barbarians and their Looney Tunes supporters,” another person wrote.
🔮 The suspect vehicle is a 2026 Land Rover that was leased by one of the world’s most celebrated composers, Harry Gregson-Williams, according to DMV records.
🔮 The family said the financial strain is particularly difficult because paperwork needed for Michelle Gigler to access insurance and retirement funds could take months to process.
🔮 “Many superintendents and school officials continue to behave as though they will not be held accountable for maintaining policies that infringe parents’ rights.
🔮 Jefferson “Zuma Jay” Wagner, a two-time former Malibu mayor and longtime surf shop owner, said he regularly witnesses dangerous encounters between drivers and cyclists on PCH, including outside Nobu.
Also by Zain Khan
Nothing else under this byline is closely related to this article, so these are simply their most recent.
All 51 articles by Zain Khan →

Topics

Black California Stanford Stanford University the Stanford Review

Subjects

Stanford PERSON · 10× Ramirez PERSON · 3× Residential and Dining Enterprises ORG · 3× Black NORP · 2× the Stanford Review ORG · 2× Billy Ramirez PERSON · 1× California GPE · 1× Stanford Residential & Dining Enterprises ORG · 1× Stanford University ORG · 1× The Post ORG · 1×

Narrative

“I was driving to school from my hometown when my friend sent me the images comparing the banner to the original photos, and I was immediately baffled,” Ramirez said. “At first, I found it hilarious that they had used AI to completely change our appearances, including my race and gender.
framing: assertive · carried by 1 article(s) · first seen 2026-09-24
🔮 Stanford’s own published guidelines say AI-generated or AI-manipulated content must be clearly identified when failing to do so could mislead audiences.

Claims (25 extracted, 2 hedged)

Stanford University has admitted it violated its own artificial-intelligence rules by using AI to alter a promotional image that transformed a male student into a Black woman — and acknowledged that the university failed to disclose the manipulation. asserted
university → admit → manipulation
The admission comes days after the student-run the Stanford Review published side-by-side images showing the original photograph and the AI-altered version. asserted
run → come → photograph
Stanford said it has now reviewed the incident and confirmed that the image was altered with AI in violation of university policy. asserted
image → say → policy
“We have reviewed the matter and confirmed that AI was used to alter the image in violation of university policy,” a Stanford spokesperson said in statement. asserted
spokesperson → review → statement
“Stanford’s AI policy is clear: The use of AI in producing or altering images of Stanford people, events, research, facilities, or achievements is strictly prohibited,” the spokesperson added. asserted
spokesperson → produce → people
“Both the alteration and lack of disclosure in this case violate that policy. asserted
alteration → violate → policy
We are working with Residential and Dining Enterprises to provide additional training and ensure proper review of all materials.” asserted
We → work → materials
The university said it has opened an investigation to learn more about the incident and “put in place protocols to prevent it from happening again.” asserted
it → say → it
“Not only the fact that the image was altered, but the ways in which it was altered, reflect a serious error in judgment,” a Stanford spokesperson told The Post. asserted
spokesperson → alter → Post
“We have apologized to the students in the photo whose images were altered or erased. asserted
images → apologize → photo
Quite simply, this was wrong and should never have happened.” asserted
this → happen → ?
Stanford’s own published guidelines say AI-generated or AI-manipulated content must be clearly identified when failing to do so could mislead audiences. uncertain
failing → publish → audiences
The university also says its communications staff should not create or publish AI-generated material that falsely depicts Stanford people, events, research, facilities or achievements. asserted
that → say → people
The controversy began after Stanford Residential & Dining Enterprises, which oversees campus housing and dining operations, used an altered version of a photograph featuring three students. asserted
which → begin → students
In the original image, Stanford junior Billy Ramirez was pictured alongside two other students holding plates of food outside a dining location. asserted
Ramirez → picture → location
But in the promotional version, Ramirez was removed and replaced by an AI-generated Black woman. asserted
Ramirez → remove → woman
The other two students also appeared to have been digitally altered, including noticeably slimmer faces. asserted
students → appear → faces
Ramirez previously told The Stanford Review that he was shocked when a friend sent him the comparison while he was traveling to campus. asserted
he → tell → campus
“I was driving to school from my hometown when my friend sent me the images comparing the banner to the original photos, and I was immediately baffled,” Ramirez said. “At first, I found it hilarious that they had used AI to completely change our appearances, including my race and gender. asserted
they → drive → race
But after looking at the comparison, I was also upset because I don’t agree with Stanford making those choices about how we were represented,” he added. asserted
he → look → choices
“Seeing my identity changed and being left out of the picture made me feel, in a way, silenced and erased from a representation that was supposed to include me. asserted
that → see → me
The school has said it is working with Residential and Dining Enterprises on additional training and review procedures following the incident. asserted
it → say → incident
“We are working with Residential and Dining Enterprises to provide additional training and ensure proper review of all materials,” they added. asserted
they → work → materials
Stanford has not publicly explained who made the alteration, what AI tool was used or why the original photograph was changed. asserted
photograph → explain → alteration
The original Stanford Dining image has also reportedly been removed from the department’s Instagram account. uncertain
image → remove → account
💬Give feedback
🕘History 🎫Support