Stanford University is facing heavy criticism after it used AI to alter the race and appearance of three students in a promotional photograph. The original image, taken in a Stanford dining hall, was digitally changed without the students' knowledge. One Hispanic student was replaced with an AI-generated Black female. The university confirmed the changes violated its own policy, removed the banners, and is now providing additional training for staff. Law professor Zachary Henderson said the incident raises serious concerns about consent and representation.
Meanwhile, the job market for computer science graduates has become increasingly difficult as companies like Microsoft, Meta and Google turn more heavily to AI for coding tasks. The unemployment rate for recent graduates sits around 7.1 percent, and enrollment in computer science programs has dropped by 8.4 percent at four-year colleges. Experts say students need to build AI and machine learning skills to stay competitive, while also developing strong critical thinking abilities.
In healthcare, a Blue Cross Blue Shield Association study found that hospitals using AI coding tools spent $942 million more for similar patient care compared to those that did not. AI tools increased the number of inpatient stays classified as medically complex by 10 percent. Researchers said the findings point to overcoding and overbilling, and they are calling for closer review of how hospitals deploy these systems.
On the security front, former CISO Pearl Almeida warned that agentic AI is creating new fears in corporate environments. Her research shows that 52.9 percent of intelligent agents in companies have no supervision or protection. The average cost of a security incident involving AI agents is $4.63 million. Almeida identified three major concerns: losing control over many autonomous agents, legal responsibility for bad decisions, and the lack of real-time monitoring tools.
Harvard Law School is also diving into the AI conversation with a new course called Debates on Frontier Artificial Intelligence Governance. The class divides perspectives into three groups: idealists who believe AI could improve the world, alarmists who fear serious harm, and skeptics who think people focus too much on future risks. Professor Jonathan Zittrain said these debates will shape laws and policies for years to come.
In other developments, Sheffield city leaders approved the use of AI to help manage city operations on September 23, 2026, marking a step toward modernizing local government services. Research from Washington State University found that AI can correct medical misinformation effectively, but tone matters. A neutral, fact-based tone works best for people who see AI as a tool, while an empathetic tone resonates with those who view AI as more humanlike. Three New York Times tech journalists also weighed in on AI, discussing the promise and risks including job loss concerns and AI safety.
Key Takeaways
- Stanford University violated its own policy by using AI to alter student photos, including changing a Hispanic student's race, and removed the promotional banners
- The computer science job market is weakening, with a 7.1 percent unemployment rate for recent graduates and an 8.4 percent enrollment drop at four-year colleges
- Microsoft, Meta and Google are increasingly using AI coding tools, contributing to fewer entry-level coding jobs
- Hospitals using AI coding tools spent $942 million more for similar patient care, with a 10 percent increase in complex inpatient classifications
- 52.9 percent of intelligent AI agents in companies have no supervision or protection, according to former CISO Pearl Almeida
- The average cost of a security incident involving AI agents is $4.63 million
- Harvard Law School launched a course on Frontier AI Governance covering idealists, alarmists, and skeptics
- Sheffield city leaders approved AI use for city operations on September 23, 2026
- Research shows AI can correct medical misinformation, but tone must match the audience's perception of AI
- OpenAI supports voluntary AI safety standards, but international agreement remains difficult as President Trump rejected a global supervisory body at the U.N.
Stanford faces criticism for using AI to alter student photos in promotional materials
Stanford University is under fire for using AI to change the race and appearance of three students in a promotional photo. The original image was taken in a Stanford dining hall. Students said the changes were unacceptable. One Hispanic student was replaced with an AI-generated Black female student. Stanford confirmed the AI use broke its own policy and removed the banners.
Stanford students protest AI altered race in university photo
Stanford University is facing backlash after using AI to alter a photo of three students for an advertisement. The image was changed to adjust race and appearance without telling anyone. Student Kavinder Patel said the school should not need to make up its diversity. The university removed the banners and admitted the AI use violated its policy.
Stanford admits AI policy violation after altering student photo
Stanford University acknowledged that it used AI to alter a photo of three students in violation of its own rules. The photo showed students in a Stanford dining hall and was used in university materials. Students said the changes spread quickly across campus. One student was replaced with an AI-generated person of a different race. The university removed the banners and is providing extra training for staff.
Stanford faces student anger over AI changed promotional photo
Stanford University is drawing criticism from students for using AI to change the race and looks of three people in a promotional photo. The original picture was taken in a Stanford dining hall. Students said the changes felt unfair and inauthentic. The university confirmed the AI use broke its policy and removed the banners. Law professor Zachary Henderson said the incident raises concerns about consent and how people are shown.
Computer science grads turn to AI skills as coding job market weakens
The coding boom is fading and computer science graduates are struggling to find jobs. Student Jack Zemke sent out over 200 applications and got only about five responses. The number of computer science graduates has grown by 50% over the past decade, but job openings have not kept up. Experts say students need to focus on AI and machine learning skills. Employers also want people who can think critically and solve problems.
Computer science graduates face tough job market as AI takes over coding roles
The job market for new computer science graduates has become difficult as AI takes over coding tasks. The unemployment rate for recent computer science graduates is around 7.1 percent. Enrollment in computer science programs has dropped by 8.4 percent at four-year colleges. Companies like Microsoft, Meta and Google are using more AI and cutting workers. Schools are now teaching students how to use AI tools and find work with smaller companies.
Sheffield city leaders approve AI use to help run city operations
Sheffield city leaders have approved the use of artificial intelligence to help with city management. The decision was announced on September 23, 2026. Leaders see AI as a tool to improve how the city operates. The approval marks a step toward using modern technology in local government services.
AI coding tools in hospitals cost Blue Cross plans $942 million more
A Blue Cross Blue Shield Association study found that hospitals using AI coding tools spent $942 million more for similar patient care. AI tools increased the number of inpatient stays classified as medically complex by 10 percent. The study looked at data from over 1,000 hospitals across the United States. Experts say this shows AI may be causing overcoding and overbilling. The study calls for closer review of how hospitals use these AI tools.
Harvard Law course explores AI idealists, alarmists and skeptics
Harvard Law School is offering a class called Debates on Frontier Artificial Intelligence Governance. The course is based on a framework that divides views on AI into three groups. Idealists believe AI could create a better world. Alarmists fear AI could cause serious harm or even threaten human existence. Skeptics think people are too focused on future fears and not enough on current problems. Professor Jonathan Zittrain teaches the course and says these debates will shape laws and policies for years to come.
Former CISO warns agentic AI is increasing security fears in companies
Pearl Almeida, a former CISO and cybersecurity expert, warned that agentic AI is making security leaders more worried. She said 52.9 percent of intelligent agents in companies have no supervision or protection. The average cost of a security incident involving AI agents is 4.63 million dollars. Almeida said traditional cybersecurity methods no longer work well with AI agents that can make decisions on their own. She identified three main fears: losing control of many AI agents, legal responsibility for bad decisions by agents, and the lack of tools to monitor these systems in real time.
AI can correct medical misinformation with the right tone
New research from Washington State University shows AI can correct medical misinformation online, but the tone matters. A neutral, fact-based tone works best for people who see AI as a tool. An empathetic tone works better for people who think AI can be more humanlike. The study found that the tone of the correction matters more than whether it comes from a human or an AI agent.
Three Times tech journalists weigh AI promise and risks
Three New York Times tech journalists discussed how they cover artificial intelligence. They addressed concerns about AI surpassing human control and causing harm. Pui-Wing Tam said the team reports on what AI does well and where it fails, like hallucinations. Mike Isaac noted growing public worry about job loss and AI safety. The journalists see their role as explaining and clarifying AI for readers.
What counts as a business moat in the age of AI
This article explores whether traditional business moats still exist in the AI era. The author argues that the ability to keep reinventing may matter more than any single advantage. Deep moats may come from judgment, culture, and the ability to learn faster than the market changes. AI may not eliminate moats but push them deeper into organizations.
AI companies and world leaders race to write safety rules
AI developers and world leaders are working on separate but similar safety efforts. Both sides agree on the need for common tests and independent review. They disagree on who should set the rules and how much power they should have. OpenAI supports voluntary standards with other frontier labs. However, international agreement is difficult, as President Donald Trump rejected any global supervisory body at the U.N. General Assembly.
Sources
- Students outraged after Stanford University uses AI to alter race, appearances in promotional photo
- Students outraged after Stanford University uses AI to alter race, appearances in promotional photo
- Students outraged after Stanford University uses AI to alter race, appearances in promotional photo
- Students outraged after Stanford University uses AI to alter race, appearances in promotional photo
- As the coding boom fades, computer science grads focus on AI skills in choppy job market
- As the coding boom fades, computer science grads focus on AI skills in choppy job market
- ‘A tool to help’: Sheffield City leaders approve usage of artificial intelligence to help with city
- Hospitals' use of AI coding tools cost BCBSA plans $942M more for similar care: analysis
- Are you an AI idealist, alarmist, or skeptic?
- Former CISO Pearl Almeida: "Agentic AI is multiplying security leaders' fears"
- AI can help correct medical misinformation—when it uses the right tone
- How Three Times Tech Journalists Make Sense of A.I.’s Promise and Peril
- What Counts as a Moat in the Age of AI?
- AI Companies and World Leaders Race to Write Safety Rules
Comments
Please log in to post a comment.