The Brief

California State University is betting that artificial intelligence will transform the workplace — and that its 470,000 students need to learn how to use it before they graduate.

Last year, CSU became the nation’s first public university system to pay for students’ access to ChatGPT Edu. It has awarded over $3 million in grants for faculty to experiment with how to use AI in classrooms. Several campuses now train students in how to responsibly use the technology.

Those measures are all pieces of CSU’s strategy to prepare students for a future labor market reshaped by AI. But the investment in AI has come under criticism from some CSU faculty, and some students remain skeptical of the value of knowing how to use AI even as most are concerned the developing technology will threaten their chances of landing a job and keeping it.

The success of CSU’s AI strategy will depend on the extent to which AI affects the workforce that today’s college students will soon enter.

Hiring is frozen in many sectors right now as employers wait to see the true extent artificial intelligence will change the labor market, said Eric Bettinger, a leading education economist with Stanford Graduate School of Education. But it’s not clear if the slowdown in the labor market can be attributed to AI.

As with previous emerging technologies, AI is forcing employers to decide whether they should replace workers with the technology, or hire more employees who can leverage AI to be more productive, Bettinger said.

New graduates are facing a difficult labor market, considering the unemployment rate for this population has steadily risen since 2022, according to the Federal Reserve Bank of New York.

“The reality is that the world outside of school isn’t slowing down, and we need to prepare our students for that reality,” James Frazee, San Diego State University’s vice president for information technology, told EdSource.

Students worry AI is taking their future jobs

Ismaeel Malik, a San Jose State University justice studies student, said he worries about AI taking over jobs in his desired field.

“Entry-level jobs are work that people don’t want to do,” Malik, who will graduate next spring, told EdSource. Malik said he doesn’t think AI proficiency will make it easier to win a position out of college.

Other San Jose State students who spoke with EdSource say they think some majors will be more affected than others. San Jose State prenursing student Jade Goodewin said the “hands-on” part of her preferred future career seems fairly AI-proof. But second-year student Vien Tang, who is studying advertising, said that AI could replace some of the work of entry-level jobs in the advertising industry.

“It’s definitely a possibility, because I think a lot of companies try to cut costs,” Tang said. “There’s a likelihood of that happening.”

Given SJSU’s proximity to Silicon Valley and many companies whose technologies are reshaping the workforce, the university hopes to be a leader in AI adoption, said Sandra Hirsh, special assistant to the university’s provost for AI initiatives.

“We need to be preparing students with whatever the latest opportunities are, but also prepare students for the fact that change and evolution are going to continue to happen,” she said.

Hirsh said companies want graduates to have a basic understanding of AI and how to use it in their work. Despite the nebulous changes happening in the job market, companies that hire SJSU graduates have told university leaders that students also need skills in leadership, inventiveness and ethics.

The CSU system is paying $13 million a year to provide ChatGPT Edu to all 470,000 CSU students.

Artur Gafurov, president of the California State Student Association and a fourth-year student at San Jose State, said many students appreciate access to AI tools like ChatGPT Edu, but don’t necessarily know how to use them responsibly and effectively.

“I know a lot of students who are taking it upon themselves to learn a bit more about the tool and restrict themselves with how they use it,” Gafurov said.

Some question the value of AI initiatives 

Other private and public universities have signed agreements with companies such as OpenAI to give students access to large language models, including ChatGPT Edu. Large public university systems, such as State University of New York, have also developed systemwide AI policies, and some colleges, such as Northwestern University, have launched AI majors.

But not everyone at CSU has embraced the growing use of AI on campuses. Some professors pushed back on the OpenAI deal after CSU signed it in early 2025. Martha Lincoln, an anthropology professor at San Francisco State University, said CSU administrators did not clearly explain how AI would be used on campuses and in classrooms by students.

“What will students be doing with AI that’s so valuable? If it’s so valuable to them professionally, educationally, where is that value and what does it look like?” Lincoln said.

Martha Kenney, a professor in the SFSU Department of Women and Gender Studies, said CSU’s claims that AI is helping prepare students for the changing labor market are dubious because the university has not clearly identified skills young people should develop for their future careers.

“It’s possible it has no use for our students in this way,” Kenney said.

According to an OpenAI spokesperson, ChatGPT Edu provides students with access to the company’s latest large language model, and includes plug-ins that students and faculty can use to personalize their accounts for teaching and learning.

University officials say providing access to tools like ChatGPT is essential to prevent a digital divide for CSU students who may not be able to afford subscriptions to these technologies.

“We believe we’re doing the right thing to serve our mission to prepare our students for the future workforce,” said Ed Clark, CSU’s chief information officer.

The future workforce

Erika McEntarfer, the former commissioner of the Bureau of Labor Statistics and a policy fellow at the Stanford Institute for Economic Policy Research, said some experts in recent years have forecast that AI would wipe out white-collar workers’ jobs. There’s little evidence that has happened, she said.

It’s difficult for universities to predict which skills will be in demand in the future, she said, but colleges should prepare their students for disruptions to their careers. She noted that roughly 60% of today’s jobs didn’t exist in 1940.

Universities need to prepare students for a tough labor market, and not just because of AI, McEntarfer said. The recent rise in the unemployment rate among new graduates started before ChatGPT was introduced in 2022, she said.

McEntarfer’s advice to recent or soon-to-be graduates is to develop skills that make you a good “human” worker, such as small talk, showing up to work on time and meeting the needs of company leaders.

“With AI you need to make sure you’re using it as a personal trainer, not a human forklift,” McEntarfer advised future college graduates. “Hopefully that’s what colleges are doing.”

Mariana Sanchez is a fourth-year journalism major at San Jose State University and a member of the EdSource California Student Journalism Corps.

This story was originally published by EdSource.

Comedian Christopher Titus brings his 11th stand-up special Doomed to Repeat to Flappers Comedy Club in Burbank on Saturday, Oct. 10.

 Titus is one of the most prolific stand-ups of his generation. Over the past 25 years, his 10 comedy specials have touched on everything from his dysfunctional upbringing in the San Fernando Valley to issues of love, marriage and parenting, and even life under the Trump administration. All summer, he's been touring his 11th special, Doomed to Repeat, which he said is about what's going on in our country right now.

"Do you remember the bicentennial? I was around for the bicentennial. Don't do the math on that," Titus said. "I remember really being proud of my country. I was a little kid and was watching the sailing ships in New York and the fireworks behind the Statue of Liberty. I was proud to be an American. And then we had our 250th birthday, and it was thrown by our drunk, abusive stepfather. It was the weirdest, worst thing I've ever seen."

Titus said he goes after President Donald Trump "pretty hard" in Doomed to Repeat, but he does it "sideways."

"I did it in Texas, and I didn't get killed," Titus said. "So, it's good."

The signature bit of Doomed to Repeat centers on immigration, a topic Titus said he has a unique perspective on, pointing out that his ancestors have been in America for 400 years.

" I thought, what an interesting perspective with all that's going on with our country right now and all the hatred and demonization of immigrants, to talk about it from a place where my family's been here before there was a country," he said.

Titus said the routine is so pointed that it's caused people who disagree with him on the issue to walk out.

"You paid. We got your drink money. We got your ticket money, and now you're leaving?" Titus said. "The bit really nails why anybody who hates immigrants is not just a hypocrite, but like a mentally ill hypocrite."

Titus said that whether it's personal or political, no topic is too messy to bring to the stand-up stage.

"If talking about my mom's mental illness and being in bars when I was four and being kidnapped ... by my dad, I mean, I don't think so," Titus said. "Pryor taught us that. Richard was raised in a brothel. I think the whole point is to drag the audience through the horror that you went through, because we're all the same machine, right?"

Although Titus is currently touring his 11th comedy special, he's already thinking about No. 12.

" I'm at 990 minutes of stand-up comedy that I've written and produced," Titus said. " Carlin's my favorite guy, him and Robin [Williams].  George did 16 specials. I'm going for 17, hopefully before I crank out."

Titus headlines Flappers Comedy Club in Burbank on Saturday, Oct.10.

Newport Beach voters might get two separate ballots for the November election, which will determine whether they vote on one local ballot measure, or four.

In one of the weirdest cases to join Orange County’s substantial history of election oddities, Newport Beach is scrambling to comply with a court order by trying to stage its own election in conjunction with the regular, general election. This is highly unusual — the vast majority of cities in Orange County, and across California, rely on the county registrar of voters to administer their elections.

Lena Shumway, Newport Beach’s city clerk, has been tasked with pulling off the flash election, including putting together voter rolls, getting ballots and election guides printed and mailed (there will be no in-person voting), and, if the election happens, counting ballots and verifying voter signatures. She told LAist she’s running into multiple roadblocks, some of them seemingly insurmountable.

And yet, she said, “We’re going to hold the election because we have to comply.”

How we got here

Newport Beach is in this bind because it challenged the validity of three ballot measures in court and, in the meantime, missed the deadline to have the Orange County Registrar of Voters place them on the general election ballot. The three measures would:

• Limit City Council members to serving a total of eight years — currently, council members are barred from serving more than eight years consecutively, but can run again as long as they take a break.
• Establish by-district voting where residents in each of the city’s seven districts vote only for their district representative on council — currently, all residents vote for representatives of each district.
• Add a “sunshine provision” to the city’s charters that would, among other things, require the city to make advisory committee meetings accessible to the public, and establish a “Sunshine Task Force” to watchdog public records requests.   
  

The measures are sponsored by the Southern California Opportunity Group, a company owned by local real estate developer Ken Picerne and his Picerne Group.

According to campaign filings, Picerne has sunk more than $1 million into these and a fourth initiative, Measure H, which will be on the regular, general election ballot. Measure H would essentially redo the city’s state-approved housing development plan and substantially decrease the number of units that could be developed.

How did Measure H make it on the general election ballot while the other three did not?

Measure H actually qualified last year, and the city promptly asked the O.C. Registrar of Voters to put it on the November 2026 general election ballot.

That didn’t stop the city from trying to block Measure H in court this year, but the effort was unsuccessful.

What’s so hard about putting on an election?

Historically, many cities did put on their own separate elections for city leaders and city-specific ballot initiatives, said Fredric Woocher, a Southern California-based election attorney. Some cities held local elections in a totally different month and day than state and federal elections. And some held them on the same day, with separate ballots.

“It was always very confusing to voters,” Woocher said. Plus, turnout for municipal elections was often dismal, leading the state to pass a law in 2015 requiring most cities to align their local elections with the statewide dates.

These days, there aren’t many companies around that offer the kinds of services Newport Beach needs, Woocher said. The city is already on its second ballot printing company, after the first company bowed out because the time frame was too tight, according to Shumway.

To help coordinate the election, Newport Beach hired a fledgling consulting firm, Stellara Group, which opened for business last year, according to its filings with the Secretary of State.

Under the contract, the city could pay up to $495,000 to Stellara and subcontractors if the election actually happens. That is still unclear.

Is Newport Beach’s parallel election really going to happen?

Shumway said there are several major hurdles that could make it impossible to successfully put on the city’s special municipal election.

For one thing, the O.C. Registrar of Voters has said it can’t provide the complete files for Newport Beach voters. This includes the signatures on file, which are used to verify mail-in voters’ identity, and confidential voter files.

Without this information, Shumway said, carrying out the election is essentially impossible “because we’re disenfranchising people … and we can’t verify signatures,” she said.

Bob Page, the O.C. Registrar, said he “felt bad” for the city during a recent media call. “I just can't put the general statewide election at risk by dropping what we should be doing to create ways to help them when their request has come way too late,” he said.

Another major hurdle for Newport Beach — the California Secretary of State has thus far refused to provide the city with the required watermark and special ballot tint. Without those two things, the printing contractor says it won’t print the ballots for the city’s parallel election.

The company has given the city until Oct. 7 to get everything in place, or else look for a new, third printing company. “At this point, it’s a waiting game,” Shumway said.

On a September evening, Adam Johnson, 40, settled on his porch chair in Morgantown, West Virginia, to go over his ballot for the upcoming midterm election.

Chatting through all the races and the different candidates, he worked to figure out who best represented his values. But rather than discussing it with a friend or family member, he held an hours-long discussion with ChatGPT — a chatbot powered by a large language model (LLM) owned by OpenAI, an AI research company.

A graphic designer studying to be an illustrator, Johnson said ChatGPT is a significant part of his life. He uses it for meal planning, assistance in job hunting and various other tasks around the house.

After prompting the AI-powered chatbot to make a list of all the races on his ballot this year, Johnson dug deeper.

Johnson is one of a growing number of voters using chatbots to research their ballots for the first midterm election since AI has taken the country by storm. About half of American adults report using AI-powered chatbots and about 42 percent of users said they use it to search for information, according to the Pew Research Center.

Several voters NPR spoke with said they rely on chatbots for down-ballot races — with thinner news coverage, to make tables comparing candidates for governors' races and to verify or debunk viral claims. Some have been using it to create tools to help others look up their own ballots. Experts say the properties that make chatbots useful also make them subtly persuasive and eager to please.

"It might select some things that would reinforce and persuade you even more towards the way that you were leaning already," said Rafael Batista, a fellow at John Hopkins University, who studies how AI shapes the way people experience the world. "So you leave more confident, without necessarily learning more about the world."

Since AI companies don't disclose how their commercial models are trained or how they choose which sources to draw on, it's hard to know what biases might shape their answers, Batista said.

Anthropic, the company that owns Claude, a chatbot, and Gemini, a Google-owned chatbot, did not respond to a request for comment.

A spokesperson for OpenAI, directed NPR to its election information and safeguards page which notes that the company continues to "monitor bias in our models to keep ChatGPT's responses politically neutral."

Concerns about AI

Johnson in West Virginia continued his research by outlining his political beliefs for ChatGPT. He stated that he usually votes for Republicans, unless those candidates are part of the Make America Great Again movement, in which case he votes for Democrats or a third party.

While deciding between candidates for the Senate race, Johnson nudged the chatbot about other issues he found important.

Johnson recognizes that AI-powered chatbots have sycophantic qualities.

"ChatGPT is just kind of a people pleaser," he said. "Whatever you lean towards, it will speak positively about it."

Still, Batista from JHU said that knowing these models are agreeable and persuasive, is not enough to "protect our minds from these influences."

Voters use chatbots in a variety of ways

Lisa Veldran, 65, was a staffer in the City Council in Madison, Wisconsin for almost four decades before retiring. This summer, she used Gemini to answer very specific questions she had about the primary election for governor.

After Sara Rodriguez, the lieutenant governor of Wisconsin, dropped out of the race, Veldran needed to choose between two candidates: Francesca Hong and David Crowley. She prompted Gemini to make a comparative table of both candidates.

Veldran said the chatbot can be a research tool, but it's not something that determines her life.

"You also have to keep in mind what's running these chatbots and AI in general," she said. "Who's paying to build these bots and what is their ultimate goal."

Lakshmi Iyer, an author living in Exton, Pa., describes herself as a power user of chatbots and thinks "it is capable of a lot of things, but the human in the loop is absolutely essential."

Iyer, 50, is part of They See Blue, an Indian American grassroots Democratic organization to drive up voter registration.

She did not use a chatbot to research her ballot, she said, but she used Claude, to create Ballot Lookup — a website that can inform Pennsylvania's voters about who is on their ballot and their stance on certain issues.

The website is described as an "independent, volunteer-built, nonpartisan lookup," and Iyer built it to help voters who may not learn more about candidates by searching across various sources.

Like other users, Iyer is aware that if chatbots "can infer your political leanings, they will push you towards what you are leaning to." To keep her own bias in check, Iyer prompted Claude to pick statements from the candidates' own speeches or from their official websites.

Some users noted that the decimation of local media has left them with fewer options for learning about down-ballot races. Tyler Black from Nashville, Tennessee, said he didn't use chatbots for House or Senate but is using them to learn about local candidates.

Black, who is a vice president at a software development company, is aware that chatbots sometimes hallucinate — that's when chatbots respond to prompts with false, misleading or fabricated answers. To overcome that problem, he often checks the information one chatbot has provided him against another.

Batista from JHU said that method of fact-checking is better than using the same model twice, but he hasn't seen evidence that it is effective.

If users are using chatbots to research their vote, Batista said, they should frame their prompts in a way that their personal preferences are hidden, and they could use incognito mode so the chatbot doesn't have much prior information on them.

Adam Johnson in West Virginia is still researching and chatting with ChatGPT about his ballot. The county commission races are still on his list to look into. But he feels like AI helped him learn everything else he needed to know for November.

"I wouldn't feel too bad if I had to do the election today," Johnson said.
Copyright 2026 NPR

The FBI arrested a California real estate agent at the Los Angeles airport as she was leaving the country and accused her of spying for China on the president of Taiwan’s family, federal officials said.

Wanying Zhang, who went by the name Heather, surveilled the Taiwanese president’s son and his family last year for the Chinese government, according to the FBI.

Zhang, 34, who lives in Irvine, was arrested Sunday at Los Angeles International Airport on charges for working for China as an unregistered agent, First Assistant U.S. Attorney Bill Essayli said. Zhang took photos and videos of her targets and sent license plate information to Chinese officials, according to a federal criminal complaint.

Zhang is a naturalized U.S. citizen, Essayli said on Fox News’ “The Big Weekend Show.” She was arrested as she was departing for a flight to China. It wasn’t immediately clear if she had an attorney who could speak on her behalf. She was due to appear Monday afternoon in federal court in Los Angeles.

The case is the latest in a series of Justice Department prosecutions aimed at rooting out alleged covert agents advancing Beijing’s interests on American soil.

Many of the cases have centered on what the Justice Department says is China’s aggressive effort to locate and harass pro-democracy activists living in the U.S. and suppress their speech. The Justice Department also says Beijing is determined to sway American public opinion in its favor by cultivating relationships with political figures seen as having access to the levers of power.

The Chinese Embassy in the U.S. did not immediately respond to a request for comment Monday and the Foreign Ministry in Beijing was closed for holidays.

Taiwan’s Presidential Office confirmed President Lai Ching-te’s relatives were the targets, without specifying if it was the president’s son, saying that “this is a classic case of transnational repression.”

“Taiwan condemns in the strongest terms such barbaric acts — which disregard human rights and national sovereignty and aim to suppress, threaten, or even harm individuals advocating for freedom and democracy or their families — as such behavior should not be tolerated in the civilized world,” spokesperson Karen Kuo said in a statement.

The criminal complaint does not name the alleged target, but the FBI’s rapid response account on the social media platform X said it was the president’s son. Ting-Yu Lai, the president’s elder son, lives and works in the Seattle area as an engineer.

At the direction of Chinese officials, Zhang and another person flew to Seattle on Sept. 1, 2025, and conducted surveillance outside the target’s home, according to the affidavit. A home security camera near the target’s home recorded the pair filming the target and their family from a vehicle, the affidavit said.

Zhang then allegedly sent a Chinese official information related to the surveillance, including photos, videos and license plate information from the target’s vehicles.

According to the affidavit, Zhang attended the University of International Relations in Beijing, which is affiliated closely with China’s intelligence apparatus. The affidavit said she and a classmate who became a Chinese government official had kept in contact since graduating around 2014, and he introduced her to another government official, and she was in contact with both in 2025.

Taiwan and China split after a civil war in 1949, but Beijing claims the island as part of its own territory. Lai, who was voted into office in 2024, has said that Taiwan and Beijing do not belong to each other and that the “People’s Republic of China has no right to represent Taiwan.”

China considers him a separatist and has called him a “destroyer of peace.”

While in office, Lai has stepped up spending on defense for Taiwan, notably pledging to increase defense spending from 3% to 5% of the GDP, and pushing through a special budget for the construction of a T-Dome air defense system.

In May, a Southern California mayor agreed to plead guilty to acting as an illegal agent for the Chinese government, and resigned from her city position. Eileen Wang, the mayor of Arcadia, had been charged with one count of acting in the U.S. as an illegal agent of a foreign government. She was accused of doing the bidding of Chinese officials, such as sharing articles favorable to Beijing, without prior notification to the U.S. government as required by law.