Tuesday, 8 September 2026

11 politicians who have changed their tune on data centers

A data center protest in Austin, Texas
A data center protest in Austin, Texas
  • The data center backlash is here.
  • Across the country, leaders are shifting how they talk about data centers.
  • Now, even previous vocal supporters are tempering their statements.

America's elected officials have a different tune when it comes to data centers.

Amid intense public pushback, elected officials and those running for office this November have moved to rein in support for data centers or enact pauses on their development entirely.

Some, like Texas Gov. Greg Abbott, represent states that are already home to massive hyperscale projects. Others, like Rep. Byron Donalds of Florida, the GOP's gubernatorial nominee, may lead states that are sorting through such proposals.

A few told us that they haven't changed their minds at all, rejecting the suggestion that they've evolved on the topic. Then there are those like Pennsylvania Gov. Josh Shapiro, who are open about how getting an earful has changed their outlook.

What is clear is that AI has noticeably changed how elected officials talk about data centers, projects that many statewide officials once clamored to bring home.

Here is how 11 officials are shifting their stances on data centers

Kathy Hochul
New York Gov. Kathy Hochul looks on
New York Gov. Kathy Hochul

In July, Gov. Kathy Hochul, a Democrat, made New York the first state to impose a one-year moratorium on what she called "hyperscale AI data centers."

The details: Her executive order states that the temporary ban applies to data centers that need 50 megawatts or more to operate. Additionally, Hochul separately ordered that data centers either pay for their own power generation or pay more for the energy they need.

Past views: In May, Hochul questioned the need for a more stringent moratorium, which was being considered by the state legislature.

"This is a local decision for municipalities, it's land use, which is the purview of local governments," she told reporters, per E&E News. "It's not a statewide approach necessarily, but it's something I'm looking at intensely."

The explanation: Hochul previously told NBC News that local communities had asked for her assistance.

"At first I thought this is more of a local decision, but then I realized they don't have the negotiating ability, the clout, the wherewithal to be able to negotiate the best benefits or to make sure that these companies, as I said in my State of the State, are either bringing their own power or paying a premium into the grid to cover the cost," she said.

Josh Shapiro
Josh Shapiro speaks during an event
Pennsylvania Gov. Josh Shapiro

In August, Pennsylvania Gov. Josh Shapiro, a Democrat, unveiled what he's called "the strictest guardrails in the nation" for future data center development.

The details: Shapiro's order, which is not a moratorium, requires developers to secure local approval before they can receive state permits, granting more power to communities that Shapiro said have been treated unfairly by some developers. Shapiro also removed data centers from fast-track permitting and established guidelines, including power generation and transparency requirements, that, if developers agree to follow, allow for permit review on a rolling basis, Spotlight PA reported.

"I've heard loud and clear from the good people of Pennsylvania as I've traveled our commonwealth, and I'm here to say today that we will not be bullied by these developers," Shapiro said.

Past views: Shapiro once openly courted data center development. When Amazon announced plans to spend $20 billion on data centers, Shapiro touted the "largest private sector investment in the history of Pennsylvania."

"With this historic announcement, we're creating opportunity for our workers, generating new revenue for our local communities, and ensuring the future of AI runs right through Pennsylvania," he said in a statement released by his office.

The explanation: When asked "what prompted his 180," Shapiro told CBS News, "Look, I listened. I think it's important for elected leaders to actually get the vibes of a community, to listen and to be able to be flexible, to be able to work with a neighborhood."

Katie Hobbs
Katie Hobbs speaks at a campaign event
Arizona Gov. Katie Hobbs

Arizona Gov. Katie Hobbs, a Democrat, pushed state lawmakers to eliminate a tax incentive for data center development that she called "a thirty-eight million dollar corporate handout."

The details: Ultimately, Hobbs and state lawmakers agreed to a three-year pause on new data center sales tax breaks.

Past views: In calling for the change, Hobbs was open about how, as a state lawmaker, she voted to create the incentive.

"More than a decade ago, we made a strategic decision to grow data centers by creating a tax exemption for them — I voted for it," she said in her State of the State address in January. "Now, Arizona is a national leader in this sector. We must ask ourselves: should taxpayers continue subsidizing the data center industry?"

The explanation: Hobbs said the incentive had boosted the industry, and now it was time for it to stand on its own.

JB Pritzker
Illinois Gov. JB Pritzker speaks during a news conference
Illinois Gov. JB Pritzker

In June, Illinois Gov. JB Pritzker, a Democrat, moved to suspend new tax incentives for data centers.

The details: Pritzker also called on the state legislature to help him advance reforms to data center development that would ensure developers pay their fair share for power use, minimize water use, and be more transparent.

Past views: Pritzker celebrated when his state reeled in major projects, such as Meta's $1 billion data center in DeKalb.

"As Illinois grows its status as a technology hub of the future, this data center is bringing investment and vitality to a newly thriving community," Pritzker said in a statement when the data center opened in 2023.

The explanation: A spokesperson for Pritzker said the conversation around data centers has changed with the advent of AI hyperscalers.

"Responsible governance means reassessing incentives when circumstances change," the spokesperson said. "He supports responsible growth of the industry, but not at the expense of higher utility bills, grid reliability or Illinois' water resources."

Wes Moore
Maryland Gov. Wes Moore speaks during a TV appearance
Maryland Gov. Wes Moore

Maryland Gov. Wes Moore said that data center projects must follow his guidelines to receive his support.

The details: In his February State of the State address, Moore said that data center projects seeking his support must hire local workers, engage the community, and cover the cost of their power needs. More recently, he has portrayed Virginia, home to Data Center Alley, as a cautionary tale.

"The way they did it there cannot and will not happen in the state of Maryland," Moore said during an event in August, according to Maryland Matters.

Past views: In 2024, Moore announced that legislation removing regulatory barriers to technology infrastructure was part of his agenda for the legislative session. When he later signed the bill into law, Moore praised what it would mean for the state.

"This bill is going to supercharge the data center industry in Maryland so we can unleash more economic potential and create more good-paying union jobs," Moore said in May 2024.

The explanation: A spokesperson for Moore did not provide a comment to Business Insider.

Greg Abbott
Greg Abbott

In August, Texas Gov. Greg Abbott, a Republican, announced an effective pause on data centers seeking to connect to the state's power grid until the Public Utility Commission and Electric Reliability Council of Texas, which operates an independent grid that services the vast majority of the state, conduct a comprehensive audit.

The details: Abbott's order cited the need to ensure ERCOT can keep up with future demand. In a statement explaining the audit, Abbott's office noted that 90% of new power requests are for data centers. Abbott also directed the two entities to gather information about proposed data centers, including their power generation, water consumption, ownership, and the extent to which the projects may need public assistance.

Past views: Abbott previously bragged about his state's ability to attract big projects. In November 2025, he celebrated Google's planned $40 billion investment by saying, "Texas is the epicenter of AI development."

The explanation: Abbott said data center developers had only themselves to blame for the backlash.

"They pop up in places that nobody ever heard of before, until they started the construction," Abbott said on ABC's "This Week" on August 23. "And so they had not been working in collaboration with the state, they'd not been in collaboration with local governments, and so they basically dug their own grave for the problem that's been caused for them, and that's why they got the backlash they deserve."

Byron Donalds
Byron Donalds speaks during an event
Byron Donalds

Florida Republican gubernatorial nominee Byron Donalds' campaign has published an ad saying he would "protect Florida families and communities from data centers."

The details: Donalds, who currently serves in the US House, has introduced legislation that would require data centers to generate their own power and to source their water from outside public systems.

Past views: According to Florida Politics, a digital media site, Donalds previously questioned other candidates who want to pause data center construction.

"There ain't no bans coming," he said.

Donalds' opponent, former Rep. David Jolly, who was previously a Republican and has since switched parties, has pledged to pause hyperscale data center construction for at least one year.

The explanation: A spokesperson for Donalds did not respond to a request for comment from Business Insider.

Spencer Cox
Utah Gov. Spencer Cox speaks during a news conference
Utah Gov. Spencer Cox

In May, Utah Gov. Spencer Cox, a Republican, issued an executive order that called for "a higher bar" for data center development.

The details: Cox's order called for state agencies to adopt a framework for considering data centers that focuses on transparency, protecting ratepayers, managing water consumption, and mitigating environmental impacts. Separately, Cox called for reforms to a state agency that pushed what became a controversial data center project.

Past views: Cox was a strong proponent of the Stratos Project, an initially proposed 40,000-acre hyperscale AI data center campus that is backed by "Shark Tank" investor Kevin O'Leary. Cox initially questioned those who criticized the proposal.

"I'm so tired of our country taking years to get stuff done. It's the dumbest thing ever. We think that taking time makes things better or safer, it absolutely does not," Cox said during an April news conference. "You get a chance to give your feedback, and then decisions get made. That's how we have to do stuff in this country and in this state."

The explanation: A spokesperson for Cox did not respond to Business Insider's request for comment. In the wake of the news conference, Cox wrote on X that his statements "did not meet the expectations I have for myself. I seek to do better." He also told reporters that public pressure helped spur his order.

"Public pressure is important, and if I had had any concerns about water and the Great Salt Lake, I would have opposed it from the very beginning," Cox said in June, per local news station KSL. "But the public has brought up some concerns that some of us didn't think about that are important, and that matters."

Jon Husted
Ohio Sen. Jon Husted emerges from a car
Ohio Sen. Jon Husted

Sen. Jon Husted of Ohio, a Republican, is sponsoring legislation that would codify elements of President Donald Trump's "Ratepayer Protection Pledge" into law.

The details: In July, Husted introduced legislation requiring states to consider standards for data centers and other large electricity users that consume 100 megawatts or more at a single site or campus. The standards include requiring those customers to cover the cost of grid upgrades needed to serve them.

Past views: As lieutenant governor, Husted celebrated Google's 2024 announcement that it would invest $2.3 billion in the state.

"Data centers are essential to our digital lives and Google's additional multibillion-dollar investment further establishes central Ohio as an important tech hub in America," Husted said in a statement at the time.

The explanation: Sen. Husted's spokesperson said his view had not changed. He sounded a similar note when Spectrum News asked him about his past statements.

"It's not really a shift. I was always trying to hold these companies accountable by making them pay local property taxes… working to make them pay their share of the infrastructure that they're required to… making sure that they hired local labor," he told the news outlet in July. "So this is very consistent for me, holding them accountable."

Vivek Ramaswamy
Vivek Ramaswamy speaks during an event
Vivek Ramaswamy

Ohio Republican gubernatorial nominee Vivek Ramaswamy has promised that if elected, he would pause approval of any new data center project announcements until state lawmakers passed his preferred requirements.

The details: In August, Ramaswamy announced he would issue an executive order to halt "the approval of any new data center project announcements" until his framework was passed into law. His requirements include providing free electricity for homeowners in communities with data centers, requiring data centers to pay full property taxes and then using the revenue to cut Ohio homeowners' property taxes, and requiring the facilities to comply with environmental standards.

"There are hundreds of data centers already across our state, and the pace of construction is accelerating," he said in a statement released by his campaign. "While I had no part in these past projects, I am committed to ensuring our policies catch up to current realities."

Past views: Previously, Ramaswamy said the fast pace of Ohio's data center buildout "was a good thing," provided the grid could keep up.

"We've got AI data centers moving to our state, that's a good thing, that's a good thing, I want the sectors of the future to come here," he said in March 2025. "But that will make a level of demand on our electric grid of a kind that we have not seen in the history of our state."

The explanation: A spokesperson for Ramaswamy did not respond to a request for comment from Business Insider.

Mike Rogers
Mike Rogers speaks during a rally
Mike Rogers

Michigan Republican senatorial nominee Mike Rogers has called for a one-year moratorium on new data centers in the state.

The details: Rogers, who represented Michigan in Congress for 14 years, told the Detroit News in August that he supported a pause but only at the state level.

"While I don't support a federal ban, Michigan needs stronger guardrails to protect community control, prevent utility price hikes, protect our water and stop pay-to-play schemes," Rogers said in a statement to the newspaper. "That is why I am calling for a one-year moratorium on new data centers until we establish a fair, transparent approval process that puts local communities first."

Past views: Previously, Rogers, who is running in one of the nation's most closely watched US Senate races, was generally supportive of data center construction, provided it had local buy-in. After the White House released Trump's AI framework in March, Rogers said, "The choice is simple: lead the charge or fall behind. Let's get this done."

In July, the former congressman told WXYZ that communities would be attracted when they see data center development could fuel property tax cuts, "I have a feeling some communities, when they see what that means to their property tax, are going to say, 'Hey, could you build another one, please real soon?'"

The explanation: A spokesperson for Rogers did not provide a comment to Business Insider.

Read the original article on Business Insider


from Business Insider https://ift.tt/LKiM6Cc

Monday, 7 September 2026

The Flock backlash is coming for retail

Ulta Beauty storefront
Retail giants like Ulta Beauty and Lowe's are being called on to end their contracts with Flock Safety.
  • Flock Safety has seen a nationwide backlash against its surveillance system technology.
  • In response, some city governments have canceled their contracts with Flock.
  • Consumers are now also calling out companies like Ulta and Lowe's, which have Flock contracts.

City governments have felt the heat to drop their contracts with Flock Safety. Now, retail giants are feeling it too.

As the backlash against Flock's security products continues, critics are targeting retail stores that have known partnerships with the company.

"There is simply no need for Ulta Beauty to be one of Flock's biggest commercial customers," Rachel LaPointe, a content creator and Substack writer, said in a video that had 732,000 views across her TikTok and Instagram as of Friday.

LaPointe, who has a Substack newsletter focused on gendered police violence, told Business Insider that her issues with Flock stem from reported instances of the surveillance system being used to track women.

"My alarm bells started ringing," LaPointe said, adding that she was surprised by Ulta's partnership with Flock, considering that a majority of the retailer's customer base is women.

An Ulta Beauty spokesperson said that Flock cameras are deployed at fewer than 1% of Ulta's 1,500-plus locations and are used to support criminal investigations, particularly into organized retail theft.

Flock has advertised its advanced surveillance system as a tool for deterring and investigating crimes, including homicides. One of its notable products includes an automated license plate reader (ALPR) that records vehicle sightings and license plates. Flock says its ALPRs do not use facial recognition and that private customers can control how long the camera data is retained and whether it's shared with law enforcement.

Flock Safety camera
Flock Safety's automated license-plate readers have been deployed by city governments and retailers.

Thousands of law enforcement agencies and communities rely on Flock's technology, according to Flock's website. The deployment of its AI cameras has also extended to large retail chains, including Lowe's and Kroger, the company wrote.

Law enforcement officers told Business Insider that Flock cameras have been instrumental in solving cases.

The technology is not perfect. A Business Insider investigation published in March showed how Flock can misread license plates, leading law enforcement to pursue the wrong person. In July, Business Insider reported that Flock cameras in the city of Roseville, California, misread license plates in 71% of alerts sent to police in 2023 and 2024.

Critics of Flock's system emerged after other publications reported law enforcement officers inappropriately using the system to track people they knew.

In Georgia, three former sheriff's deputies were arrested this August on charges that they used Flock's license plate data to track people with whom they had personal relationships, The New York Times reported.

404 Media reported in May that Texas deputies tapped Flock's database during a probe involving a woman suspected of carrying out a self-administered abortion.

A Flock spokesperson did not respond to a request for comment.

Reports of abuse of Flock's system, along with broader anxieties about privacy and mass surveillance, have prompted city governments to reexamine their contracts with Flock and spurred individual efforts to take down the cameras or block them with trash bags.

Florida's Department of Transportation announced on August 31 that it would ban automated license-plate readers from state roads. DeFlock, an open-source license plate reader tracking website, said that it has recorded 153 cities rejecting, canceling, or deactivating their Flock agreements.

A DeFlock spokesperson did not respond to a request for comment.

'Mass Surveillance Isn't Pretty'

Ultraviolet, a nonprofit women's advocacy group, put out a petition on August 26, titled "Hey Ulta: Mass Surveillance Isn't Pretty," calling on the beauty retail giant to end its contract with Flock, which the group sees as a threat to women's privacy and safety.

"While the majority of Flock ALPR cameras are deployed via contracts with law enforcement agencies — which is where much of the abuse we're concerned about takes place — retailers like Ulta are also playing a key role in normalizing AI surveillance and feeding the personal data of everyday people into a vast network that could be accessed by all kinds of actors that don't always have our best interest in mind," Jenna Sherman, a campaign director for Ultraviolet, told Business Insider.

Sherman said that Ulta's answer — that Flock is used to combat retail theft — doesn't justify tracking every single inbound and outbound customer.

"As we'd say in the South, which is where I'm from, Ulta is using a shotgun to kill a fly. And we're all getting shot," she said. "The vast majority of people driving past these cameras aren't criminals — we're customers, employees, parents, and people running errands. Yet Flock and other ALPRs track our license plates and movements anyway, often without our knowledge or consent."

Sherman said the petition received 13,650 signatures as of Friday afternoon.

Flock Safety camera
Flock cameras have been vandalized amid criticisms against the surveillance tech.

Fight for the Future, a nonprofit that advocates for digital rights, launched a campaign called "Deflock Lowe's" in February, providing a platform for people to organize protests and sign a petition demanding that the home-improvement retailer "drop Flock."

Reem Suleiman, a campaign director for Fight for the Future, told Business Insider that the initiative began in part after 404 Media publicized records showing that a local Texas sheriff's office could access Flock's license plate readers at 173 Lowe's locations.

"We were outraged to learn that Lowe's was further endangering the immigrants and day laborers, who were already being targeted by ICE at stores, by participating in this surveillance network and by voluntarily sharing camera data with other law enforcement agencies," Suleiman said.

Suleiman said there have been about 10 events organized and that the petition has received over 2,200 signatures. Lowe's had not responded to the organization since the initiative began, Suleiman said.

A spokesperson for Lowe's did not respond to a request for comment.

For LaPointe, there's a bit of a personal tie to Ulta, growing up as a girl in the Midwest, where strip malls with Ulta stores were common sights. She said shopping can now be "frustrating" when a company's practices conflict with personal values.

"Nowadays, I feel like there are so many moral conflicts with a lot of stores, and people have to navigate what it means to them to shop ethically," she said. "And it's definitely impacted my relationship with Ulta, having shopped there myself plenty over the years."

Read the original article on Business Insider


from Business Insider https://ift.tt/e7UxTM8

Sunday, 6 September 2026

They were getting closer to student-loan forgiveness. Then their progress was stripped away.

Photo collage with paperwork, a stressed person with a notebook, payment amounts, and a graduation cap silhouette, surrounded by Private lending elements.
Some student-loan borrowers lost payments toward the Public Service Loan Forgiveness program.
  • Public Service Loan Forgiveness promises student-debt relief after 10 years of qualifying payments.
  • The Trump administration recently removed some student-loan borrowers' PSLF payments, citing coding errors under Biden.
  • "My life has basically been hijacked," one borrower said.

Christina Quinones was two payments away from reaching the threshold for student-loan forgiveness.

She'd spent nine years and 10 months pursuing Public Service Loan Forgiveness, which forgives student debt for government and nonprofit workers after 10 years of qualifying payments. Quinones, who works in a Texas school district, was looking forward to the relief she believed was coming soon.

Then she got a letter from her servicer this summer: her 118 qualifying payments were decreased to 98, tacking on nearly two more years to her path to loan forgiveness.

"I was just shy of two payments," the 42-year-old said. "What happened?"

The answer lies in the changes the federal government made to student-loan accounts in recent years.

In May 2024, the Biden administration extended a credit for some payments that had been previously deemed ineligible for forgiveness through PSLF or income-driven repayment plans.

Now, the Department of Education says coding errors "resulted in inaccurate payment counts for some borrowers." A department spokesperson did not comment on the number of borrowers affected by the "code errors" and had PSLF payments removed.

Business Insider reviewed the letter from Quinones' servicer, which said that "one or more of your qualifying payments has changed status to 'non-qualifying,'" and that in order to receive credit, she had to switch to a qualifying repayment plan.

Since President Donald Trump's student-loan repayment overhaul took effect July 1, borrowers have reported a range of challenges with their payments, including higher bills, inaccurate payments, and difficulty getting clear answers from their servicers. The Education Department said its changes are intended to curb excessive borrowing and simplify a complex repayment system, but borrowers said they're struggling to navigate the changes while planning for their financial futures.

Quinones is feeling defeated. She said her servicer could not give a clear explanation of why her payments were deemed ineligible, and she's holding off on making further ones until she receives clarity on her PSLF status.

"I was super excited because I was so close to relief," Quinones said. "And now I'm just so upset."

Holding out for student-loan forgiveness

Jennifer Krabill, 47, is in a similar situation. She spent more than a decade working for the states of New Jersey and Maryland, and made 90 qualifying payments toward PSLF. In August, her servicer said eight of those payments were deemed ineligible.

"My life has basically been hijacked for another eight months," Krabill said.

Her $120,000 student-loan balance comes from a joint bachelor's and master's program at American University and a J.D. from Delaware School of Law. She's spent more than seven years making monthly payments, and it hasn't been easy: In 2016, her wages were garnished when she couldn't make her student loan payments on a five-figure income.

She said PSLF was a beacon of hope; loan forgiveness would give her the freedom to move to Ohio, to be closer to her family. With the payment timeline now extended, she feels like she's "in a hostage situation."

"It makes me physically sick," Krabill said. "I don't even know what I rightfully owe because of all this extra stuff they've tacked on."

Borrower advocates have criticized the Trump administration's move to rescind PSLF credits. Julie Margetta Morgan, an Education Department official under Biden, said in a statement that "someone needs to stand up for borrowers here."

"Make sure public servants get what they're owed. Make servicers pay for errors they caused. Make Federal Student Aid accountable for the problems it caused, too," Morgan said. "Do all of that before you even think about reinstating a single loan."

The PSLF program is also embroiled in litigation. The Education Department planned to roll out a rule in July that would have limited eligibility for the program, excluding employers that did not align with the administration's definition of "public service," such as those that supported gender-affirming care. A federal judge blocked the rule from being implemented, and the Trump administration recently appealed the ruling.

Krabill's current plan: send letters to her local lawmakers and the Consumer Financial Protection Bureau regarding her situation, in hopes of regaining the PSLF credit she lost.

"I'm beyond frustrated," she said. "It took me a couple of days to even just recover. I can't believe months just got taken."

Have a story to share about student loans? Reach out to this reporter at asheffey@businessinsider.com.

Read the original article on Business Insider


from Business Insider https://ift.tt/9dtJ5xH

I lost my sight 5 years ago. Tesla's Cybercab gave me a glimpse of regaining my independence.

David Moss and Gary Moss
Gary Moss, right, who is blind, took his first ride in a driverless Tesla Cybercab with his son, David Moss.
  • Tesla launched its two-door, steering-wheel-free Cybercab on Thursday.
  • Gary Moss and his son, David Moss, were among the first public riders to try the new vehicle.
  • The father, who is blind, said a car like the Cybercab could help him regain independence.

This as-told-to essay is based on a conversation with Gary Moss, 62, a resident of Puyallup, Washington, and his son, David, a Tesla owner. Moss, who is blind, attended Tesla's Cybercab launch event with his son in Austin. Tesla made the Cybercab available to the Austin public following the event. The two hailed several rides in the Cybercab. The conversation has been edited for length and clarity.

Gary Moss: Before I lost my vision, driving gave me the freedom to go where I needed, when I needed.

I used to drive 30,000 to 40,000 miles a year for my business. There was nothing more exciting to me than going to a place that I'd never been before on the open freeway. I found that to be as relaxing as a 10-mile run. It was almost the same euphoria.

Almost five years ago, around the end of October, I lost sight in one eye. Within a three to four-month period, my second eye went bad. I have cancer that's treatable but not curable.

David Moss: They realized he had a softball-sized tumor in his brain.

Gary: I'm now pretty good with a cane. I can get around my house without a second thought, and I'll do a lot of walking around the cul-de-sac. But there are a lot of places I can't go.

If I'm sitting in the middle of the day and I want to go to my favorite coffee shop, I used to have the freedom to do so. Now, I have to plan trips and bum favors. People are very nice to take me places, but I still have to ask them, and they have to be willing and have the time and ability to take me where I want to go.

David: I mostly help with driving him around or running errands. I'm also his personal stylist and shopper, too.

We used to go to a lot more sporting events and big things in heavy-traffic areas with lots of people around. We don't do that nearly as much anymore.

The Cybercab feels like a 'living room'

Gary: The first thing I noticed about the Cybercab was how easy it was to get in. I appreciate the doors opening automatically. It was like having the chauffeur open the door and say, "Oh, Mr. Moss, glad to see you."

Tesla's regular flush handles can be a little difficult for me. You have to hit them with a thumb and grab them at the same time. With the Cybercab, they open up, you step in, and you step out. I'll put my hand on the door so I don't run into the edge.

Gary Moss and David Moss
Gary Moss and his son, David Moss, took several rides in a Cybercab after Tesla held a launch event in Austin.

David: If you walk toward the doors in a way that they'll be in your path, the doors will only open part of the way.

Gary: Tesla also had a little phrase: "We want you to feel like you're in your living room." That really worked especially for me because I'm not distracted by things going by. I'm sitting there in a very comfortable seat listening to good music and chatting with my son.

At the end of my time in my second living room, all of a sudden, I show up where I want to be.

Other people's driving scares me, but these cars don't.

David: He's taken six Cybercab rides by now.

Gary: When I'm with people who are gracious enough to drive me someplace, I try not to criticize too much. Countless things grab my attention, like making last-minute decisions to change course or hitting the brakes. That virtually never happens when I'm in a Tesla.

'It's not all the way there'

David: On a busy street with five or six Cybercabs on the block, there weren't any audible cues that could help my dad know which one was ours. That's one of the things that would need to improve.

Gary: When you're in a spot with several Cybercabs, there needs to be a way to make it easy for me to know which one I'm walking up to.

David: We appreciated that we could ask Grok, the AI voice assistant, a few things: where we were, how many miles we had left, and the time. But my dad wouldn't be able to change or set the destination.

There was more voice-activation stuff than we anticipated, but it's not all the way there for someone like my dad.

Cybercab console screen
The Cybercab's interior user interface mainly consists of a large screen.

Gary: One of the problems with ride-hailing apps is that they require the ability to use the phone, and I can only use my phone with voice activation.

If I could just say, "Hey Siri or Grok, I need a cab to take me to the grocery store or take me someplace to eat," and in a couple of minutes one showed up, and all I have to do is touch a button and I'm gone — I can't believe the options that I'd have. Sometimes I'd like to go someplace and get something to eat, and nobody wants to take me there.

And instead of a brief tone, announcing that we're here, maybe it would also help if the Cybercab gave me a brief summary of where we're parked and, using the cameras, what it's going to take to walk the, say, hundred feet between where we're parked and the door of the building I'm going into.

David: A lot of these things would be irrelevant if there were personal ownership of a Cybercab, if my dad only had to walk out to the driveway, and he could use Grok to control the car.

Gary: I'd buy a Cybercab. I'd feel pretty lucky. The experience is so close to being everything I need right now.

Read the original article on Business Insider


from Business Insider https://ift.tt/vmCUEFo

Saturday, 5 September 2026

Two scientists made huge bets on 'impossible' cancer treatments. They're finally paying off.

Tal Zaks
Tal Zaks, the former chief medical officer of Moderna, said he took "a very large financial bet just based on first principle beliefs."
  • 2026 is the year of the cancer breakthrough, with innovations from Revolution Medicines and Moderna.
  • "I'm almost tired of saying paradigm shift, but I haven't figured out a new phrase," one expert said.
  • Innovators Greg Verdine and Tal Zaks pushed boundaries with risky, groundbreaking experiments.

Short sellers love to profit from biotech's big fails. The category remains their favorite to bet against because roughly two-thirds of publicly traded biotech companies ultimately go broke one way or another, losing money for investors.

This summer, however, that strategy went bust. Biotech has experienced unusual success with some of the category's biggest, longest cancer moonshots coming to fruition.

On August 19, Moderna announced positive late-stage results from a melanoma trial, sending its stock surging 177% — up to $174 per share — on the promise of the very first personalized cancer vaccine. A week later, on August 26, another astounding development: The FDA approved a pill to help treat late-stage prostate cancer. The drug is a new molecular "glue" from Revolution Medicines, a Silicon Valley-based biotech company, and it's capable of drugging so-called undruggable cancer targets.

"This is a whole new era," Dr. Andrew Coveler, who directs the Fred Hutch Pancreatic Cancer Specialty Clinic in Seattle, said. "I'm almost tired of saying paradigm shift, but I haven't figured out a new phrase."

These medicines are both used in concert with existing drugs, and neither is considered a cure for cancer. Instead, they're seen as breakthrough options that open up a new playing field for fighting cancer, targeting different mutations, and training the immune system in new ways.

The moves prompted a huge surge of cash for investors in the companies, prompting celebration at corporate offices.

"We raised our glass of champagne, but now back to work," Dr. Alan Sandler, the chief development officer at Revolution Medicines, said.

The successes weren't always obvious, though.

Two innovators — a chemical biologist and an oncologist — had both the vision and the capital to finally make these groundbreaking ideas work, despite many failures before.

A chemist who raised $125 million to chase an idea inspired by nature

greg verdine
Greg Verdine is a chemical biologist and former Harvard professor turned biotech entrepreneur. He turned to nature for inspiration to design a new molecular "glue" to target cancer.

In 2012, after four years of pleading for cash, Greg Verdine finally scrounged up the investors to give him $125 million for a cancer drug.

He didn't have a drug yet. In fact, he wasn't exactly sure how it could be made. He just had an idea inspired by nature for targeting one of cancer's most "undruggable" mutations, RAS.

The problem with RAS is that it is so smooth. Unlike other molecules, this protein has no characteristic sticky pockets or holes for a drug to glom onto. Even if you could figure out how to attach something to the smooth surface of this protein, you'd run the risk of having your therapy stick onto everything else in the body, too, potentially damaging healthy organs.

"People told me it was crazy. It'll never work"

- Greg Verdine

This is why most chemists before him gave up and called RAS an impossible mutation to target, even though it often drives tumor growth in some very tough-to-treat cancers, including pancreatic cancer, colon cancer, and lung cancer.

"I just found it totally unacceptable, pathetic, that we would say, 'Oh, well!' You've got all these patients," Verdine said. "We know exactly what's driving their cancer, but 'Oh, that's undruggable.' Come on!"

Verdine knew that molecular glues already existed in the natural world. Compounds produced by bacteria and fungi cause proteins to bind together, altering their function without completely destroying them. Some immunosuppressive drugs for organ transplantation take advantage of that biology.

The remaining question was: How can you engineer a new way to do that, targeting the "undruggable" RAS cancer mutation? By sticking the drug to another target, then gluing it onto RAS.

"People told me it was crazy. It'll never work," Verdine told Business Insider. "It's the combination of a target that's never been drugged before, and drugging it in a way that's never been done before. That sounds like compounding risks."

It turns out, Verdine is a pretty persuasive guy. He flew to France, where investors at Sanofi, a large French pharmaceutical company, were intrigued by his bullishness and agreed to pay 50% of his Series A funding round for his startup Warp Drive Bio.

"There was no IP. There was nothing. It was just an idea that we could drug RAS using molecular glue technology, and I had no idea how you might do it," Verdine said.

In 2018, the technology Verdine pioneered at Warp Drive was acquired by Revolution Medicines, where the chemistry was finessed further. Fourteen years after his initial bet, the FDA approved a drug by Revolution last week that does exactly what Verdine imagined.

ben sasse
Former Sen. Ben Sasse has late-stage, metastatic pancreatic cancer. He's been taking the new daily pill and says it helps.

Former Nebraska Republican Sen. Ben Sasse, who has stage 4 pancreatic cancer, calls Revolution's pill, which he's been taking in a clinical trial, a "miracle" worker for reducing both his pain and dramatically shrinking tumor volume.

"I have much, much less pain than I had four months ago, when I was diagnosed," Sasse told "60 Minutes" in April. "And I have a massive 76% reduction in tumor volume."

For now, the drug, called Rasonque, is chiefly recommended for patients with advanced cancer and is meant to be used after chemotherapy. It costs about $40,000 a month, before insurance, and side effects include rashes, mouth sores, and some diarrhea.

Taking a $450 million bet based on a belief

tal zaks
Zaks was confident a cancer vaccine could work. It was just a matter of timing it right for the immune system.

For cancer doctor Tal Zaks, former chief medical officer at Moderna, the ambition was similar.

He knew a cancer vaccine could work. It just required the right architecture, asking the right questions, and conducting the right trial at the right time.

Oh, yes, and also lots of cash, including a $450 million deal with Keytruda-maker Merck to co-develop the technology.

"Where the naysayers are right in this regard is that for cancer vaccines, neither mice data nor phase 1 data even are indicative of whether it's going to work, so you're really taking a very large financial bet just based on first principle beliefs," Zaks said.

"The truth is that from a business perspective, this one was tough — it was a high risk, high reward type bet — it required a significant investment of capital."

He knew lots of cancer vaccines had failed before. He was confident there was a way to win, though, if the immune system was called up in the right way and at the right time.

"From a business perspective, this one was tough — it was a high risk, high reward type bet"

- Tal Zaks

Moderna's melanoma vaccine harnesses the genetic makeup of each patient's tumor — teaching the body to attack its own unique cancer mutations. It's administered after a patient's initial treatment and is designed to reduce the risk of cancer recurrence.

"That's really the success of Moderna. I was very adamant when we started this at Moderna that we either funded to the end of a randomized phase 2 or it's not worth doing," he said, referring to the mid-stage clinical trial where researchers tested their vaccine against standard methods for preventing relapse.

The results were a "wow," Zaks said, with Moderna's vaccine almost cutting in half the risk of skin cancer recurring. "This is not about curing mice or showing cool data for a scientific publication. This is about getting conclusive proof."

Moderna's new vaccine for melanoma patients is not FDA-approved, but could be on shelves as soon as 2027, pending a positive review. Common side effects include things you'd expect for a vaccine: some fatigue, chills, and mild pain at the injection site.

But there are caveats to this medication, too. It has to be taken in concert with Keytruda, a $12,000 to $25,000 IV immunotherapy. It also has to be used after a person's tumor has already been surgically removed.

40 years of discovery science also helped

moderna lab worker
A personalized cancer vaccine takes only a few weeks to produce, thanks to advances in genetic sequencing, and mRNA vaccine technology.

Of course, there were other discoveries along the way that helped shepherd these breakthroughs, including a far better understanding of cancer genetics, a few key technical solutions, and additional cancer drugs that work in concert with the new technologies. In short, decades of plodding, hard work by scientists across the country and around the world.

"It's taken at least 40 years," mRNA vaccine researcher Phil Felgner, a University of California, Irvine professor who helped develop early versions of the lipid nanoparticle technology that powers the new vaccines, said. "It's taken this much time for it to mature."

And the work isn't over. Revolution is testing a pill for lung cancer, and seeing how its drugs might compete head-to-head against chemotherapy. Moderna is testing vaccines for lung, bladder, and kidney cancers.

These are "approaches that, hopefully, over time will have less toxic side effects to patients, and more targeted effects on the cancer," Dr. Christopher Flowers, division head of cancer medicine at The University of Texas MD Anderson Cancer Center, said.

All accomplished with some new glue and a shot.

"Ideas sound simple once you get it out, and then there's elegance in that," Coveler, the pancreatic cancer doctor, said. "But I think the design and the specificity of these things was in no way easy to figure out."

Read the original article on Business Insider


from Business Insider https://ift.tt/85ayb07

Hikers 'relied too much' on AI and had to be rescued. A ranger says it wasn't the first AI-related mishap he's seen.

A hiker walks along a rocky mountain trail beneath a blue sky with wispy clouds.
Three hikers rescued from Mount Shasta said they'd relied too much on AI, the Forest Service said.
  • Three hikers rescued from California's Mount Shasta on Monday said they "relied too much" on AI.
  • Ranger Nick Meyers said it was the third time this year that AI played a role in hiker mishaps.
  • One woman, he said, ignored a "no camping" sign because AI told her it was a good place to camp.

Nick Meyers, lead climbing ranger for the US Forest Service on Mount Shasta in California, has seen people make plenty of mistakes in the mountains over the past 20-plus years. But he's seeing a new trend in 2026: hikers erring with AI's help.

Meyers, who is also the director of the Mount Shasta Avalanche Center, said there have been at least three mishaps this year in which hikers or campers were led astray by AI or relied too heavily on the tech.

The most consequential came last weekend, when three 20-year-old men got lost and had to be rescued after summiting Mount Shasta, California's fifth-highest peak, located in the Cascade Range at the northern end of the state.

The "novice climbers" had set out on Saturday on the Clear Creek route and established base camp at a prohibited site, the US Forest Service said. They woke up at 3 a.m. on Sunday and set out for the summit, equipped with enough food and water for an eight-hour hike. They summited 16 hours later at 7 p.m.

While trying to descend the 14,179-foot peak in the dark, the climbers went off route, and one of them injured his knee. The cellphone they'd been using for navigation died. They eventually had to stop and form a makeshift camp to spend the night, without adequate food, water, or equipment, the Forest Service said. At 9:30 a.m. on Monday, climbing rangers reached the party and helped them descend.

The group told the rangers they had used Google Gemini, YouTube, and the AllTrails app to plan their climb, authorities said.

"We relied too much on AI rather than our own critical thinking," the group told rangers, the Forest Service said.

Meyers said AI was not the "nail in the coffin" for the group, but that they had made a series of mistakes that all added up. He said the roughly eight-hour estimate they got from AI wasn't generally wrong, but that the group hiked far slower than that, hadn't packed enough provisions for the extra time, and kept going after blowing past the recommended turnaround time of around noon.

Still, he said the group itself believed AI contributed to their mistakes and that there may be too much reliance on the technology, especially among younger hikers.

Google did not respond to a request for comment. The company says on its website that Gemini can make mistakes and hallucinate, and that users should double-check its responses.

Three hikers with helmets and backpacks traverse a rocky high mountain slope beneath rugged peaks and patches of snow.
The hikers rescued on Mount Shasta said they used Google Gemini to plan their trip.

AI-related mishaps are on the rise

Meyers said rangers encountered two other people this summer who ran into issues after using AI to plan their trip.

He said one woman had set up camp in an area where camping was not allowed, as was clearly labeled at the site. When rangers informed her she couldn't camp there, she said, according to Meyers, "Well, ChatGPT told me I could camp here."

"We're like, 'Well, the sign says no camping,'" Meyers said.

In another incident, a man was talking to rangers about a hiking loop he was planning to take, but they were confused by what he was saying. When they asked exactly what he was planning, he pulled out a AI-generated map of the mountain and his route.

"The map was terrible, totally terrible," Meyers said. "It had features that were on complete wrong sides of the mountain."

OpenAI says on its website that ChatGPT can make mistakes and might respond confidently "even when it's wrong." The company also lists practical tips for how to use its chatbot responsibly, including: "Use ChatGPT as a first draft, not a final source. Please check important information."

People making mistakes in the mountains is nothing new, and rangers have long advised hikers to do their homework, plan carefully, and consult official sources, such as the Forest Service or the National Park Service. AI can be used as one part of that planning, too.

"AI can be a great tool," Meyers said. "But it's not the only tool in the toolbox."

AI is especially dangerous to rely on in the mountains, where you're dealing with Mother Nature, unpredictable weather, and risky terrain. Meyers said hikers also need to rely on what he called "old school skills" — like talking to a real person, a ranger or an experienced climber, and getting information from firsthand sources.

Mount Shasta specifically has a website updated with the latest information, and climbers are encouraged to call to speak with a ranger and ask questions. AI can also be used to find the real people you need to call.

"Our website will pop up," he said. "We've queried it."

The reality is that safety in the mountains is dependent on a series of small decisions you make over and over again as you're climbing, Meyers said, and part of that is based on "mountain sense" built up through experience.

"One thing we don't ever want to do is discourage people from adventures," Meyers said. "We just want to make sure they plan properly and encourage folks to start at the shallow end of the pool."

His advice: Find a good mentor. Talk to real people. Visit the ranger station.

"You can do your online research," he said, but also "fact-check yourself a little bit and use those critical thinking skills."

Do you have a story to share about using AI in the outdoors? Contact this reporter at kvlamis@businessinsider.com.

Read the original article on Business Insider


from Business Insider https://ift.tt/OXwFC0o

Friday, 4 September 2026

How opposing camps of parents pushed Mamdani to weigh in on AI in school

Ilana Glazer speaking into a megaphone at an anti-AI rally in New York City in June.
Ilana Glazer speaking into a megaphone at an anti-AI rally in New York City in June.
  • Mamdani paused most AI use in schools this week for at least a year.
  • The decision came after parents on both sides of the issue lobbied his government for months.
  • One side argued that AI is bad for a child's development. The other: It's necessary for a career.

Craig Garrett had no idea his 5-year-old daughter was talking to what he described as an AI chatbot at school every day.

That was until the Brooklyn-based book editor attended a meeting held by his school district, where he said educators revealed the extent of AI use in his daughter's classroom.

"I was kind of horrified and shocked and disappointed," Garrett told Business Insider. "The whole notion of a 5-year-old plugged into a tablet with their headphones on and a microphone, and speaking to this animated character that's talking to them like it's their friend is just deeply wrong."

That meeting, which took place in May last year, "radicalized" many of the parents in the room, Garrett said. It led him to join the AIM coalition, a group of New York City parents opposed to AI in the classroom. AIM stands for AI Moratorium.

Those concerned parents are often the loudest voices in the room, but they aren't the only concerned parents, and they aren't the only voices. There is a cohort on the other side that believes students preparing to enter the workforce need to be taught to use AI, or else be left behind.

"In the last six months alone, the references to having AI skills in job postings in the United States have more than doubled," Keri Rodrigues, the founding president of the National Parents Union, which has a chapter in New York City and which supports AI use in schools, told Business Insider. "Parents want their kids to understand how to utilize this technology."

These two groups of parents, representing opposite sides of the debate, engaged in a monthslong effort to lobby Mayor Zohran Mamdani's administration, which this week ultimately sided — at least for now — with the anti-AI contingent.

Mamdani announced on Wednesday that his administration would impose a one-year moratorium on student-facing AI platforms in New York City public schools for students in elementary and middle school. Under the new restrictions, Amira, the program Garrett's daughter was using, will have its AI features turned off.

A spokesperson for New York City Public Schools said in a statement that the school system "engaged with various stakeholders — educators, students, parents, advocates, union partners, experts, and other community members" to shape the policy.

The AIM coalition had two meetings with NYCPS Chancellor Kevin Samuels in recent months, members said.

Parents in the AIM coalition called the moratorium a victory, though it fell short of their demand to pause the tech for two years. They described feeling blindsided by the extent of AI use in the classroom. Some said they feared that data about their children would be scraped by AI companies. Others said they didn't want chatbots taking the place of teachers.

"It really started from concerned parents finding out things about what their children were experiencing and being completely appalled," said Liat Olenick, a leader of the group. "It really took off because I think it struck a nerve."

Olenick recalled multiple incidents of parents approaching her in bewilderment about the AI programs their children were using. "All of a sudden, one day, people were like, 'Wait, my 7-year-old's using AI? I never gave permission for that,'" she said.

Olenick said the group began campaigning for a moratorium earlier this year. She said the group's lobbying efforts included conversations with staff members in the mayor's office, as well as the meetings with Samuels.

"The organizing really made a difference," she said. "We changed the conversation."

Members of the AIM coalition pictured after a meeting with Chancellor Samuels at the Tweed Courthouse in May.
Members of the AIM coalition pictured after a meeting with New York City Public Schools Chancellor Kevin Samuels at the Tweed Courthouse in May.

Under the new AI guidelines unveiled on Wednesday, teachers will still have access to generative AI tools, which they can use to plan lessons but not to grade assignments. High school students will participate in "AI literacy" courses and have limited access to five pilot AI programs, none of which are chatbots.

Joe Philleo, the CEO of Edia, one of the programs permitted for high school use, applauded the one-year pause in a post on X. "Mamdani's plan strikes a reasonable balance for this school year," he wrote. "Young kids should not be glued to their screens all day."

The CEO of Amira, meanwhile, defended the software's role as a tool for teachers.

"We understand how challenging it is for districts to ensure that they're putting appropriate AI safeguards in place, and appreciate that, as New York does that hard work, they're also taking care to ensure that teachers who are using Amira to diagnose reading challenges and improve literacy outcomes for kids can still access the tools that are working for them," the company's CEO, Mark Angel, said in a statement.

Mamdani described the moratorium as the most "comprehensive of its kind anywhere in the country." It comes during a nationwide public reckoning over how best to expose children to generative AI, as educators bemoan how students are using the technology to cheat on assignments and parents demand more regulation.

Rodrigues, however, said the ban was a "hysterical response" that could leave students unprepared for a job market that values AI skills more than ever.

Though Rodrigues said AI has its place in the classroom, particularly when combined with strong and consistent oversight by human teachers, she stressed that there are certain functions — like grading, counseling, and imposing discipline — where AI has no place.

The new policy, she said, uses a sledgehammer instead of a scalpel.

"You are seeing a knee-jerk political response to something that is complex, requires nuance, requires foundational understanding, and should be evidence and data-driven," she said.

City officials emphasized during their press conference announcing the moratorium on Wednesday that the pause was temporary as they assess how AI affects a child's ability to learn.

Kaliris Salas-Ramirez, a neuroscientist and associate professor at Rutgers University, and a mother of two sons in New York City's public school system who supports the pause, said research suggests young children using AI tools risk preventing their brains from developing problem-solving skills.

"We're actually not allowing the brain to consolidate that information in a comprehensive way," she said, referring to the use of AI during a young person's formative years.

For Garrett, whose daughter is now entering second grade, he's relieved the Amira app won't be used as often.

"It's not like this app was doing something that teachers couldn't do before," he said.

Read the original article on Business Insider


from Business Insider https://ift.tt/vAhbTuV

11 politicians who have changed their tune on data centers

A data center protest in Austin, Texas Jay Janner/The Austin American-Statesman via Getty Images The data center backlash is here. Across th...