NEW YORK (AP) — OpenAI received a subpoena from several states as part of a probe into the safety of users of its chatbot as it prepares to offer stock to the public for the first time.

The company behind the popular chatbot, ChatGPT, said it will respond to the inquiry “constructively” and that it already has in place measures to protect its customers.

“AI is a new and powerful technology, and we work every day to safely bring its benefits to people in a responsible way,” an emailed statement from a spokesperson said. “We take the concerns raised by state attorneys general seriously.”

OpenAI has drawn criticism for ChatGPT allegedly offering encouraging words to users thinking of killing themselves or engaging criminal acts. It also has come under scrutiny for how its uses health data and other personal information of its customers.

On Thursday, the company was sued by a Canadian blaming the chatbot for her daughter’s decision to die by suicide. Earlier in June, the Florida attorney general after two separate shootings where alleged gunmen were reported to have asked ChatGPT questions while planning their crimes.

OpenAI said in a statement that its models repeatedly encouraged the individuals to seek real-world support, including from mental health professionals. The company also said it has cooperated with law enforcement in both shooting cases.

The new probe comes just a few day after with U.S. security regulators for a highly anticipated initial public offering of stock. Artificial intelligence rival SpaceX on Friday. The rocket maker founded by Elon Musk also runs an AI business responsible for a rival chatbot called Grok.

How governments should respond to the potential for good and possible dangerous of AI is becoming a big political issue.

Regulators Europe into Musk’s Grok over antisemitic content and sexualized material, include . And another chatbot company preparing an IPO, Anthropic, was directed by the Trump administration Friday to to users abroad for national security reasons.

The OpenAI subpoena was earlier reported by The Wall Street Journal.

The Associated Press sent emails to a dozen state attorneys general Saturday asking for details of the probe but has not received any responses.

In its statement, OpenAI highlighted measures it has taken to keep children using its chatbot safe.

“Today’s ChatGPT includes a more protective experience for minors and people experiencing difficult situations, with safeguards that direct them to real-world resources and trusted human contacts,” the statement read in part. “We believe kids should be treated like kids, which is why we built age prediction, released parental tools to guide their children’s use of AI, and disallowed advertising that targets kids.”

What most people might consider good news — the creation of 172,000 jobs in May from a U.S. economy that was more robust than expected — was treated by Wall Street as a horrible turn of events.

So horrible that it justified wiping out $2.3 trillion in stock equity.

The broad selloff on Friday was the stock market’s worst showing since October, led by big declines in technology, especially semiconductors, A.I. and social media.

Granted, valuations were beyond stretched and priced for perfection in the tech sector, including an assumption of lower capital costs in the years ahead.

The realization that might not happen sent investors running for the hills.

Colorado’s hardest hit company was Ascent Solar Technologies, which lost 17.5% in value. The solar research firm has a very small market capitalization, and its shares are highly volatile.

More informative is the $1 billion hit that Riot Platforms, based in Castle Rock, took. It is an industry leader in Bitcoin mining and digital infrastructure for crypto firms. Its shares were down 10.2%.

It followed the downward move in Bitcoin, which has lost about 15% of its value this week and dropped below the psychologically important $60,000 mark on Friday for the first time since late 2024.

Bitcoin, the world’s leading cryptocurrency, managed to recover to $61,000 at the close on Friday. It is down by more than half from its all-time peak of $126,000 reached last October.

Although the biggest losses were centered on tech companies like Nvidia, which lost $150 billion in market value, energy and mining companies didn’t provide a haven.

Shares of Denver-based Liberty Energy, which provides oilfield services, were down 9.85% in value, even with concerns mounting that the conflict with Iran could turn hot again.

Gold historically represents an asset that can serve as a store of value when inflation and currency devaluations are threatening. Despite that rising uncertainty, gold mining companies took it on the chin on Friday.

Shares of Denver-based Newmont Corp., the world’s largest gold mining firm, dropped nearly 8% in value, shaving off $9.2 billion in value.

Newmont alone accounted for more than half the lost value among Colorado stocks on Friday, according to a Denver Post analysis.

Other mining and royalty companies with big losses included SSR Mining, Vista Gold, Royal Gold and Solitario Resources.

A few consumer-facing companies, led by food providers, managed to eke out some of the largest gains on an otherwise rough day.

Natural Grocers, Pilgrim’s Pride and The Simply Good Foods Co. managed gains ranging from 3.4% to 2.1%.

The Liberty family of companies, which are hedged from higher interest rates because of their strong cash flows, stayed in the green.

Frontier Group Holdings Inc., parent of Frontier Airlines, also defied the odds with a 1% gain.

One explanation is that consumers might be drawn more to ultra-low-cost carriers to save money in a world of rising inflation.

But then again, if they are fully employed and wages are rising, the odd they can keep the 70% of the economy they represent chugging along will improve.

Re: “Polis vetoes ‘surveillance pricing’ bill,” June 3 news story

Gov. Jared Polis’ veto of the anti-surveillance pricing bill proves once again he’s just a Republican wearing a liberal costume. His excuse? The bill, which would have banned companies from using AI and “big data” to manipulate prices and wages based on your personal circumstances, might “interfere with the free functioning of markets.”

Sure. Because nothing says “free market” like corporate algorithms tracking your every vulnerability to extract maximum blood from your stone. This isn’t capitalism; it����ap corporate sharecropping scaled to the state level.

Consider this hypothetical: A software engineer gets laid off but has some savings. On the way to a job interview, he pulls up to a gas pump. Between inserting his card and pumping, the oil company runs an instant “wealth check.” Seeing his healthy savings balance, the algorithm spikes his price per gallon.

He arrives at the interview. Instead of offering a salary based on market value and experience, the employer scrapes data on his time out of work, his dwindling savings, and his chronic illness. They craft an offer just high enough to keep him from drowning, complete with a health plan that conveniently excludes his condition.

Is this the “free functioning of markets?” No. It����ap an asymmetric data war where citizens are completely outgunned. But hey, as long as Gov. Polis can keep defending the “freedom” of monopolies to pickpocket your data, who cares about the actual people?

For someone who just moved from Florida in part to escape this nonsense, Polis disappoints.

Tom Gawronski, Evergreen

Climate crisis is front-page news

Re: “U.N.: Next five years could smash temperature records,” May 29 news story

Banging the climate crisis drum: Last Friday, The Denver Post relegated a major U.N. climate report to page 12. Ho-hum, the world scientific community keeps banging that old drum about the climate. No big deal. We haven’t gone off the cliff — yet.

But there is a cliff there. Scientists just don’t know when the edge — the tipping point — will be reached.

Have you noticed all the floods, droughts and temperature records we are experiencing (again) this year? Are you concerned about this being a really bad fire year? Drill, baby drill continues as President Trump says we have to produce more oil, while the report concludes that oil and gas is the major contributor to the issue. Ho-hum.

As a committed climate activist, I plan to keep banging that old drum and supporting the rapid transition away from oil and gas to renewable energy.

Marc Alston, Denver

Sarah Woodson for House District 42

Sarah Woodson is a breath of fresh air for the residents of House District 42. A new voice of reason and common sense for everyday Aurorans stressed out by politicians on the far right and left who only support special agendas, not their constituents.

It was 40 years ago that Aurorans trusted another homegrown centrist political newcomer who went door to door to listen to his neighbors and represent them, not the special interest lobbyists that swarm over our Capitol like the miller moths and locusts of summer.

It takes a strong voice from a future leader like Sarah Woodson who listens first to the people and serves them, and not the special political insects. Too many people are again suffering real economic hardships, like how to simply pay for this week’s groceries, while the politicians of the far right and left play off one another and do nothing to help the common people.

It is time again to support a homegrown political newcomer who will serve us, the people, not them, the special interests.

Steve Ruddick, Aurora

Editor’s note: Ruddick is a former Aurora state representative.

To send a letter to the editor about this article, submit online or check out our guidelines for how to submit by email or mail.

LOS ANGELES (AP) — Shortly after issued his calling for the robust regulation of artificial intelligence, the Instagram meme account Saint Hoax posted a video to its more than 3 million followers about the pope’s call to “disarm” AI. “Love my woke pope (I’m not even Catholic),” the caption read.

In another viral post, one X user referenced a common meme in response to the encyclical, writing: “The atheism leaving my body the moment the pope starts talking about how AI is an affront to God and the new Tower of Babel.”

That kind of reaction to Leo’s first encyclical, “Magnifica Humanitas” (Magnificent Humanity), has been prevalent online since its release last week. That enthusiasm is driven in part by a perception, especially among young people, that few political or global leaders grasp or take seriously the known and potential ramifications of AI’s rapid rise. Leaders often have worked to accommodate the tech industry, citing the need for economic growth — and along the way, critics say, cozying up to wealthy CEOs.

“People have really been looking for a response to AI,” said Isabel Thurston, a 27-year-old comedian from Boston. “This was the first — at least in my sphere of the world — world leader to make an announcement to this magnitude.”

History’s first U.S.-born pope has demonstrated a willingness to embrace aspects of contemporary culture. He was recently spotted wearing Nike sneakers under his vestments, and in his encyclical, Leo quoted the wise wizard Gandalf from the “Lord of the Rings,” a series by Catholic author J.R.R. Tolkien.

“It����ap clear that this is written by an American pope. There’s a spirit breathing through this document of an emphasis on individual freedom, on human happiness and human dignity,” Robert Orsi, a professor of religious studies and history at Northwestern University, said of Leo’s encyclical. “At times, I thought the language really resonates with the Declaration of Independence.”

This specific kind of cultural fluency may help explain some of his viral moments as the leader of the ancient faith.

Just weeks earlier, a group of youths visiting the Vatican coaxed the 70-year-old pontiff to do a viral hand gesture on camera known as the 6-7 meme — a meaningless “brain rot” joke among young people. Though it����ap clear in the video that the pope, like most adults, doesn’t understand what they’re asking him to do or its significance, he does it anyway and is met with enthusiastic cheers. A week later, he did it again while smiling and waving to crowds from the popemobile.

The portrait that has emerged from these instances is one of playfulness, but also intentionality. The pontiff reiterates throughout “Magnifica Humanitas” that it is the church’s responsibility to engage contemporary questions and challenges.

“Her mission has a historical scope and entails a responsibility for the way in which social relations are built,” Leo wrote about the Catholic Church. “She cannot consider herself a stranger to the forces shaping society. On the contrary, the Church actively participates in the processes by which society grows and is organized.”

Since his election last year, Leo has made a point to directly converse with — and sometimes critique — various aspects of society, ranging from politics to entertainment and sports.

Pope Francis, Vatican II and other preludes

Orsi studies the relationship between Catholicism and modernity, which he says have often historically been at odds with one another. He said Leo’s encyclical and his broader papacy, like that of his predecessor Pope Francis, is informed heavily by the still-polarizing Second Vatican Council, which brought modernizing reforms to the church more than 60 years ago.

“It����ap speaking with a Vatican II voice to the modern world. So, it����ap not a voice of condemnation, but it����ap a voice of respect,” Orsi said of Leo’s encyclical. “Pope Francis, in a sense, was the necessary prelude to this kind of encyclical. I think Francis gave a really strong encouragement to take a clear critical voice on these urgent questions.”

That����ap not to say there haven’t been detractors to Leo’s approach. Some criticized his decision to present his encyclical co-founder Christopher Olah. The Vatican decided to involve the tech company as part of its decade-long effort to engage Silicon Valley in dialogue over the human cost of AI.

In the roughly 42,300-word document, the pontiff exhorts all “men and women of goodwill” to not be afraid to get their “hands dirty on the ‘construction site’ of our time.”

That willingness has sometimes led to measured but very public rebukes of policies, actions and leaders, including U.S. President Donald Trump and the ongoing war in Iran. Some conservatives including Vice President JD Vance, who is a Catholic convert, have invoked the concept of “just war” theory in response to Leo’s critiques.

Church teaching has long allowed for “just wars” — the use of force to stop an unjust aggression — as long as certain conditions are met. But Leo directly addressed this doctrine in his encyclical, calling it “outdated.” “Humanity possesses far more effective and capable tools for promoting human life and resolving conflicts, such as dialogue, diplomacy and forgiveness,” he wrote.

Hollywood to baseball: American culture at the Vatican

Last November, the pontiff hosted a “World of Cinema” day at the Vatican with actors and filmmakers including Cate Blanchett, Viggo Mortensen, Gus Van Sant and Spike Lee, who gifted Leo a custom New York Knicks jersey with the number 14 and the name Pope Leo on the back.

“Cultural facilities, such as cinemas and theaters, are the beating hearts of our communities because they contribute to making them more human,” Leo told his Hollywood audience. “The logic of algorithms tends to repeat what ‘works,’ but art opens up what is possible.”

He has also not been shy about his love of the Chicago White Sox, sometimes sporting baseball caps or posing with bats — the latter inspiring a kind of subgenre of Leo memes. “POV: you’re a priest who just asked ChatGPT to write your Sunday homily,” the Rev. Harrison Ayre posted on X with a photo of Leo holding a bat and smiling.

Shortly after “Magnifica Humanitas” was released, Thurston, who is Catholic, posted a video of her and a friend drinking margaritas while meticulously studying and discussing printed pages of the encyclical. It has racked up more than 3 million views on Instagram.

“An aspect that made the video going viral really joyful for me was to represent all of the Catholics or lapsed Catholics or adjacent interested parties as really celebrating what Pope Leo is saying in his encyclical,” she said.

Orsi said this strategy is coming at a crucial time for the Catholic Church following years of reckoning with its legacy of . “I think a lot of people who moved away from the church are now saying, ‘Wait, maybe the church does have something to say to the modern world,’” he said.

Associated Press religion coverage receives support through the AP’s with The Conversation US, with funding from Lilly Endowment Inc. The AP is solely responsible for this content.

The measure, which passed as , focuses on safeguarding chatbots’ interactions with children and teenagers. It faced controversy while it was under consideration in the state legislature, in part because some advocates for tight regulations thought the final version of the bill didn’t go far enough. It ultimately passed with bipartisan lineups in support and opposition.

Starting Jan. 1, the new law requires that chatbots regularly announce to users who are minors that they are operating on AI. Developers of chatbots will also be required to implement measures aimed at preventing emotional dependence, and they will be tasked with putting reasonable safeguards in place to prevent chatbots from generating sexually explicit content.

The chatbots must refer users to a crisis services provider if they exhibit suicidal ideation.

“The unfortunate reality is that AI chatbots have encouraged suicide attempts and engaged in romantic interactions with minors,” said Rep. Sean Camacho, a Denver Democrat in a statement. “Our new law protects users, especially children, from misleading AI chatbot conversations.”

Sen. Iman Jodeh, an Aurora Democrat, said in a statement that the law “takes the first step toward establishing commonsense guardrails so that our children are encouraged to turn to trusted adults, not to AI chatbots, in times of need.”

Earlier this month, Sen. Kyle Mullica, a Thornton Democrat, invoked the death of a 13-year-old girl, Juliana Peralta, whose mother said she was sexually groomed by a chatbot and that the chatbot did not do anything to intervene when the girl shared suicidal thoughts. Juliana later died by suicide.

Her mother, Cynthia Montoya, who lives in Mullica’s district, argued the bill was insufficient. She cited that it would set a “technically feasible” standard for incorporating the new protections in chatbots, and she said that would give tech companies too much deference. She called for Polis to veto the bill.

And for many of the state’s executives, balancing the tremendous power and possibilities of AI against its potential drawbacks is an ever-increasing concern.

New data from the found that AI is transforming job functions faster than most organizations can adapt.

In last year’s survey, 78% of respondents said generative AI, agentic AI and machine learning will have the biggest impact on workplaces over the next three to five years. That percentage jumped to 93% this year.

The capacity AI has to remake a company’s workforce is already playing out.

Earlier this year, Denver-based Angi Inc. cut 350 jobs as part of an AI-driven efficiency effort, while Block tied to automation.

More broadly, Meta, the parent company of Facebook and Instagram, this week laid off 8,000 workers as the company remakes itself with AI.

The speed at which changes occur in the state might be stemmed, in part, by a significant gap between AI ambition and AI execution, the survey highlighted.

Only 7% of organizations report enterprise-wide adoption. Half remain in the pilot phase, 33% describe themselves as scaling across multiple functions, and 10% are either in early exploration or have not begun using AI at all.

The roundtable’s survey draws on executives from across more than a dozen industries and organizations ranging from under 100 Colorado-based employees to more than 5,000, with both statewide and Front Range operations represented.

“Colorado employers are not waiting for the future of work to arrive, they’re retooling how they hire, where they invest and how they partner with organizations. They’re moving faster than ever before,” said Eyal Darmon, who leads Accenture’s public service AI practice, addressing business leaders Tuesday at the United Club Lounge at Empower Field at Mile High during a workforce summit hosted by COBRT.

“The real question that this room has to grapple with today is whether the rest of the ecosystem, policymakers, education, workforce development – frankly, all of us – how can we all keep up with that pace?” Darmon said. Executives from Colorado companies and tech giants such as Nvidia and Hewlett Packard Enterprise discussed how employers are adapting to an AI-driven economy that is increasingly becoming the new normal.

“The biggest mistake I see companies make is waiting till they have it all figured out — every day you’re waiting, you’re falling behind. Every day you’re waiting is a day of exponentially less productivity,” said Sean Young, Nvidia’s Director of Enterprise Industry Marketing. More than use Nvidia AI technologies, along with 15,000 global startups in Nvidia Inception.

The company, known for inventing the graphics processing unit, is a leader in AI innovation, including what it describes as the for building quantum processors.

Young told attendees he views ChatGPT and other large language models as relatively low risk, and said companies should make AI tools widely available so employees can experiment and figure out how to use them.

“If you take a step back and look at this from an academic perspective, 500 years from now, this is a historical moment that we’re living in,” Young said.

“Yes, jobs are going to change, the world is going to change, but this is your opportunity to do something about it.”

Darmon said at Accenture, AI has transformed both how the company operates internally and the nature of what it delivers to clients.

“I think I sat on these couches last year and told you that you should think of AI as having 1,000 interns supporting you. Today, it’s 1,000 Ph.D. students that can go down into a level of capability that you’ve never fathomed before, and it’s not slowing down,” Darmon said.

He also said “durable” human skills, such as relationship-building, will be increasingly important for success, as technical skills become more widely accessible.

When employers were asked in the survey what skill gaps are most difficult to fill, 90% pointed to durable, human skills such as critical thinking, communication, and adaptability. That far outpaced digital and technical skills (43%), senior leadership pipeline (47%), and skilled trades (30%).

More specialized technical categories rated slightly lower, including AI literacy and applied AI skills, industry-specific certifications, business and financial acumen, and company- or role-specific skills.

The survey said this does not mean technical skills are unimportant, but that employers value the capacity to learn, adapt and collaborate at least as much as any specific technical competency.

Despite AI’s growing impact on work, the Colorado Business Roundtable report said employers are not bracing for mass displacement.

When asked how AI will affect workforce demand in their industry, 40% respondents said they expect no significant change in workforce size, and another 17% expect moderate or significant increases. Only 3% anticipated a significant decrease, while the rest expected modest reductions or remained uncertain.

Still, that outlook contrasts with developments over the past year, as companies in Colorado and across the country have already carried out layoffs while restructuring operations around the technology.

Companies including Atlassian, Crypto.com, HP, IBM, Salesforce, Pinterest, Dow, UKG and Snap, among others, have connected to AI initiatives.

Major companies like Microsoft, Meta and Google are not only adopting AI, but have also set of its use among its employees. In one case, a mid-sized consulting and accounting firm has even made employee bonuses dependent on .

These policies have further fueled skepticism about AI’s growing role in the workplace.

For Jeremy Bray, a former software developer who recently lost his job, the increasing use of AI across the industry has disrupted his expectations of a stable career path in tech.

“There hasn’t been nearly the amount of jobs that there used to be,” he said, adding that many companies over hired during and after the COVID-19 pandemic before later scaling back and shifting resources toward AI.

“I’ve been applying to 300 jobs a week and I hardly ever hear back from any of them — and then with people getting laid off from the bigger companies by the tens of thousands, there’s just so much competition for what few jobs there are. It����ap making it really hard.”

When asked about the uncertainty about how AI could reshape jobs, Hewlett Packard Enterprise Innovation Architect Baradji Diallo said the picture that emerges is one of transformation rather than elimination.

Yet Diallo said he realized when it comes to AI, most problems are not about the technology itself. The problem, he said, is adoption.

“It shouldn’t be human versus AI, it should be humans actually changing how we work because of AI,” Diallo said.

“We have to ask ourselves, ‘what is AI actually enabling us to do?,’ because reality is, it will replace some of those low-level tasks that are repetitive.”

Bray shared a similar view. Online customer service and call center roles could likely be among the first jobs to be displaced.

He said many of the tasks performed in those positions can already be handled by , systems that autonomously performs tasks by designing workflows with available tools.

With AI agents now across a range of industries and applications, including customer support, appointment scheduling, data analysis, software coding, healthcare administration and financial services, Bray said he began to recognize how significantly AI could shape the future of his career.

“If you can just tell a bunch of agent processes to go and build something, there’s not a whole lot for you to do other than wrangle those processes and oversee what they do,” he said.

The roundtable’s survey said 73% of respondents are not tracking productivity or performance outcomes tied to their AI investments. Among those that are, the most common measures include cost reduction, time savings and customer experience, while fewer track metrics such as revenue per employee or error reduction.

Only 13% cited AI and data talent gaps as a primary obstacle, suggesting the bottleneck is organizational readiness, not workforce supply.

Yet as concerns like Bray’s become more common, the survey found companies are beginning to adjust how they structure work around AI.

Employers are rethinking which tasks belong to humans, which belong to machines, and how the two should interact. The most common organizational response is redesigning work processes and job roles (60% of respondents), followed by combining automation with human augmentation (47%). Workforce reduction was cited by just 13%.

“I think a lot of (companies) don’t seem to understand how imperfect a lot of the systems are,” Bray said.

Bray, who has worked in the tech industry for more than a decade, said he has used ChatGPT, Microsoft Copilot and Claude for coding, data analysis and creative work.

“Working with the systems, I find all the time where they can give bad results. They aren’t consistent by nature. They don’t give the same results every time when you ask the same question. There’s no really good way of quality control,” he said.

Paul Boyer, a psychotherapist for Kaiser Permanente in Oakland, California, is experiencing the AI revolution firsthand. He’s a little underwhelmed.

The health giant has rolled out a new suite of note-taking software, made by healthcare AI pioneer Abridge, intended to summarize a patient����ap visit at supersonic speed. For many clinicians, the technology soothes one of the persistent headaches of their lives — administration and paperwork.

But the AI scribe caused another headache for Boyer and his colleagues: It is “not super useful.” They end up correcting the computer-written notes.

Abridge is “not good at picking up on clinical nuance, at picking up on the emotional tone” that can be critical in the mental health field, Boyer said. For example, for manic patients, what����ap said is less important than how it����ap said, Boyer said, and the software struggles with picking up on those cues.

Note-taking software isn’t the wave of the future; it����ap the wave of the present. Hospitals nationwide are implementing it. And researchers are finding some benefits. A year after installation, doctors who used these products the most saved more than half an hour of work daily, according to published in April in the Journal of the American Medical Association.

Many doctors love the products where they’re deployed — several to the scribes.

Nevertheless, as Boyer’s example shows, there are persistent questions about the systems’ quality. While Boyer and his colleagues spend time correcting notes, safety researchers worry clinicians might not be diligent about catching errors. That might mean future doctors rely on bad information.

Abridge says it evaluates its scribes at every stage of deployment, including with head-to-head tests against previous versions of the software.

“Following deployment of a model, we monitor clinician edits, star ratings, and free-text feedback from clinician users about note quality,” the company’s director of applied science, Davis Liang, told KFF Health News in a statement.

Artificially intelligent scribe software is part of a swarm of AI-powered tools coming to healthcare. Clinicians and patient-safety advocates say government regulations are not well constructed to guard against the threat that the new technology will miss or obscure important details of patients’ conditions, potentially harming them.

“There is currently no safeguard in place” to vet scribe software at the federal level, said Raj Ratwani, a researcher specializing in human factors — that is, how people interact with technology — at MedStar Health, a large hospital system based in Columbia, Maryland.

Ratwani worries that safeguards on health software will relax even further. from the Office of the National Coordinator for Health IT — the body that regulates electronic health records, the central chronicle of care for patients — could weaken requirements to make medical records understandable, easy to use, and transparent about the use of AI, Ratwani said. And an incomprehensible record could confuse clinicians and lead to errors.

Beginning in the Obama administration, the Health and Human Services Department����ap IT office , in which developers try their products on doctors and nurses. Regulators also sought to require more transparency from companies in the surging market in AI tools.

Both of those requirements are axed in the proposed rules from HHS Secretary Robert F. Kennedy Jr.’s health IT office.

Doctors and other health practitioners consult records for clinical information, such as scribe notes summarizing the history of patient care and lists of drugs and therapies their patients have used. Doctors also input orders for care.

Poor or cluttered design of a records system “might make the list of medications so complicated and confusing that the ordering provider selects the wrong medication,” Ratwani said.

Abridge’s general counsel, Tim Hwang, said the company “broadly supports” the government����ap rules as a “necessary modernization” that “accommodates the speed at which AI is evolving.”

The old rules “put way too much burden” on electronic health record systems, said Ryan Howells, a principal at Leavitt Partners, which consults for digital health companies. Leavitt supports the proposals.

Dropping requirements, the administration argues, will result in more innovation and competition. The electronic health record market has steadily consolidated, with hospitals and other clinicians choosing from fewer vendors.

A 2022 study found the top two vendors, Epic and Oracle Health, of the hospital market. And Howells argued too many rules burdened providers looking for good record systems. Federal regulations, Howells said, are “the single biggest inhibitor to true clinical innovation.”

The Trump administration proposal to remove requirements governing records is overbroad, some critics say. It removes regulations intended to keep records secure. It also eliminates privacy protections for sensitive medical data they safeguard, overhauls standards governing the formats data is sent in, and more. The rule may give clinicians “more health IT choices to meet their needs through increased competition,” the government wrote in its proposal.

HHS’ health IT office declined comment, noting the proposal is still winding through the regulatory process. Public comment closed in February.

But most concerning to some — even in the hospital and developer sectors — are proposals to scotch prerequisites to ensure new products are tested on actual users, and to ensure AI tech’s decisions are transparent to doctors and nurses.

“Historically, hospitals and health systems have been challenged by the black box nature of certain AI tools and how the algorithms are developed,” the American Hospital Association’s Jennifer Holloman said. And with more AI tools flooding the market, the association , transparency is even more critical.

Complaints about the safety of electronic health records are long-standing, even for seemingly straightforward tasks. Ratwani likes the example of ordering medication for a given condition.

“The physician is trying to order Tylenol, and the medication list can be so confusing that there’s 30 different versions of Tylenol all at a different dose and for different purposes, when in reality that could be designed much more simply and make it easier for the physician to actually pick the right type of Tylenol that they’re ordering,” he said.

Real-world user testing was intended to simplify record design for doctors. But the administration is ending that requirement in a confusing way, said Leigh Burchell, vice president for policy and public affairs at Altera Digital Health, an EHR developer.

In Burchell’s interpretation of the rules, which refer to “enforcement discretion,” a principle in which the government can opt not to enforce certain rules, companies are still required to do the testing — the part that takes work — but are not mandated to report their results to the feds.

The administration is also ending a Biden-era idea to create AI transparency “model cards.” The concept was that clinicians could explore the data used to train AI tools that advise clinicians with a simple mouse click. But few took advantage of the year-old tool, President Donald Trump’s regulators say.

Still, hospitals and doctors are wary of removing it. The tool “provides information on how a predictive or generative AI application was designed, developed, tested, evaluated and should be used. These data are critical to foster trust in AI tools and ensure patient safety,” the AHA wrote in a comment letter to the HHS IT office. The American College of Physicians , saying a “lack of clarity could undermine clinician trust, increase liability expense, and erode the patient-physician relationship.”

Even developers aren’t totally sure about the idea. Burchell said the electronic health records trade group she’s part of had “a lot of different perspectives” on the issue. “Normally, we tend to be a bit more aligned on our responses.”

Still, Burchell’s group thought companies should be transparent about the data AI relies on to make decisions and how it comes up with recommendations.

Evidence for AI tools’ effectiveness or contradictory.

A comparing 11 AI scribes for potential use as a pilot in the Veterans Health Administration found the software performed worse than humans across five simulated scenarios. “Although ambient AI scribes can generate complete notes, the overall quality remains broadly below that of human-authored documentation,” the authors noted, with the omission of information being particularly concerning, given the potential to affect follow-up care.

The vendors in the VA study weren’t identified, for what the authors called “contractual reasons.”

And that����ap just one type of AI tool. A wave of them is coming, each needing its own evaluation, to say nothing of tools that have already been installed.

Boyer said he can mostly ignore his AI scribe, for the moment. But he worries that management will design his job around the expected time savings and schedule more patients — meaning he’d need to spend more time both with patients and correcting the software’s errors.

A KP spokesperson, Vincent Staupe, said the company does not require its clinicians to use AI.

“When I am correcting that note, I feel like this is too much work,” Boyer said. “This is definitely making this worse, and this is taking up time that I need to not be spending on correcting an AI tool.”

©2026 KFF Health News. Distributed by Tribune Content Agency, LLC.

Google will soon unleash a wealth of new artificial intelligence-powered tools and systems, including an AI assistant that will help users by proactively performing tasks on their behalf.

, the recent buzzword of choice for tech firms, was a central focus of Google’s annual developers conference, Google I/O. The upcoming AI agent, Gemini Spark, was one of many of the company’s announcements from the conference Tuesday.

“We are firmly in our agentic Gemini era,” Google CEO Sundar Pichai said Tuesday before a packed amphitheater near the company’s Mountain View, California, headquarters. “I’ve played around with all sorts of agents and you can really see the potential, but it����ap still early days when it comes to making agents easy to use, super secure and truly helpful.”

Google and its corporate parent, , have poured billions into AI development. Its top finance executive said on a call with investors in late April that this year’s capital expenditures may climb as high as $190 billion. But the investment seems to be paying off, with its quarterly earnings showing . The stock has climbed another 11% since the report last month.

Pichai said during the keynote address that the Gemini app had 400 million monthly active users last year, but that usership has now surpassed 900 million, more than doubling in a year.

The latest version of Gemini is here

Google’s latest family of models, Gemini 3.5, is rolling out Tuesday to billions of global users beginning with Gemini 3.5 Flash. The Flash model is focused on speed, and Google says 3.5 Flash is its strongest agentic and coding model yet, but it����ap also about four times faster than some competitors.

This model is now the default for the Gemini app and “AI mode” on Google search. The company is also working on the 3.5 version of Gemini Pro, which it says it����ap using internally and expects to launch next month.

Gemini 3.5 was developed with new, more advanced safety training and mitigations, meaning its models are less likely to generate harmful content or to mistakenly refuse to answer safe queries, the company said.

Google also announced a new model, Gemini Omni, which will enable users to create high-quality video by making a query with any input, be it text, images, videos and audio. The video Omni creates can then be edited easily though a conversation with the model. Users will eventually be able to create images and audio with Omni, but there were no details about when those features will be rolled out.

The company said Omni’s videos will appear more realistic than videos created by other models because of its understanding of forces like gravity, kinetic energy and fluid dynamics.

Gemini Omni Flash, the first of the Omni family, is launching Tuesday for Google Al Plus, Pro and Ultra subscribers through the Gemini app and Google Flow. Beginning this week, it will be available at no cost on YouTube Shorts and YouTube Create App.

All videos created with Omni will include Google’s imperceptible digital watermark, SynthID, but Google is also adding content credentials verification to the Gemini app. This tool determines if content like photo or video was created by AI or captured with a phone camera and edited with AI tools. It will be available in search in Chrome in the coming months. Google also announced AI companies Open AI, Kakao and Eleven Labs are adopting its SynthID technology to more of their AI-generated content.

A 24/7 agent and wearable AI

Powered by Gemini 3.5, Gemini Spark will be able to complete mundane, routine tasks like sorting through meeting notes, emails and chats and then creating a document with the biggest takeaways and to-dos. Unlike other available agents, Spark is based in the cloud, so it continues working in the background even when users shut their laptops or lock their phones.

The proactive nature of AI agents is what differentiates them from chatbots, and that has also led to some anxieties about the technology’s power. Gemini Spark is designed to ask for permission before performing “high-stakes” tasks like sending an email or making a purchase, the company said.

Select testers will have access to the agent beginning Tuesday, and the company plans to roll out the beta mode to U.S.-based subscribers to its Google AI Ultra tier.

Later this summer, Gemini Spark will operate directly within Chrome, the company said.

Among the many AI-centric announcements at the conference was an update on the long-awaited smart glasses from Google, of which there will be two kinds: audio glasses that offer spoken help in your ear, and display glasses that provide information visually. The audio glasses will come first, with the company expecting them to arrive later this fall. Users will be able to say “Hey Google” or tap the side of the frame to access Gemini, which will then assist with navigation, managing communication on their phone, real-time translations and other tasks.

Google partnered with Samsung and eyewear brands Gentle Monster and Warby Parker to create the glasses and gave the first look at two designs on Tuesday, with sunglasses from Gentle Monster and glasses from Warby Parker. Those designs will launch as part of the eyewear brands’ full collections later this year, Google said.

More AI in search and shopping

At last year’s conference, the most talked-about development was the introduction and rollout of . The feature gives users a more conversational answer to their query before providing relevant links, building on previously implemented changed how users experience and interact with the platform.

AI mode queries have more than doubled every quarter since its launch last year, and the tool recently surpassed 1 billion monthly users, according to Liz Reid, Google’s head of search.

The new default model in search will now be Gemini 3.5 Flash and the company is introducing what it calls an intelligent search box. This change, which Reid says is the biggest upgrade to the search box in 25 years, means the box will adapt to accommodate longer queries and it can help users write out their questions with AI-powered suggestions instead of traditional autocomplete.

Users can also search using multiple modalities, using text, images, video, files and even Chrome tabs as search inputs. The new search box is starting its roll out Tuesday in all countries and languages where AI mode is currently available.

The company also announced a new tool, the Universal Cart, which it called “a truly intelligent shopping cart.” It works across merchants and across services so users can add things to their cart while browsing Google search, chatting with Gemini, watching YouTube, or reading emails in Gmail. The cart then runs on Gemini models to go to work as soon as an item is placed in the cart, looking for deals and price drops, providing price history information and alerting users when something comes back in stock.

The Universal Cart tool will be available to users on search and the Gemini app this summer, with YouTube and Gmail to follow.

Associated Press Writer Barbara Ortutay in Oakland, California contributed to this story.

In the coming weeks, the council plans to convene a group of community advocates and experts to make recommendations on what those regulations should be.

Data centers are buildings where tech companies, including giants such as Amazon and Netflix, store their servers and storage systems. The centers make everyday services like email, online banking, telehealth, streaming services and artificial intelligence language models possible for Denverites. They can also suck up massive amounts of water and energy.

The facilities have become a hot-button issue across the country as communities push back against construction in their areas. Neighbors of the centers complain that they create constant noise, pollute the air with diesel generators and exacerbate water shortages.

On Tuesday, the Jefferson County commissioners imposed a moratorium on new data centers there, specifically freezing any new applications for rezoning or development of the high-tech facilities for 10 months.

In Denver, dozens of residents signed up to speak Monday night during the public hearing portion of the City Council meeting. Many of them came from northern Denver, where neighbors have expressed frustration about a new CoreSite data center there. A community group from the Globeville and Elyria-Swansea Neighborhoods, located near CoreSite, hosted a press conference ahead of the meeting.

“This is an important first step,” Alfonso Espino said. “But we want to ensure … that we will make this moratorium mean something.”

There are 50 data centers in Denver, said Councilwoman Serena Gonzales-Gutierrez, one of the sponsors of the moratorium.

About half of them — at least in terms of square footage — are in the council’s District 10, which includes central neighborhoods like Golden Triangle, Union Station and Congress Park, said Councilman Chris Hinds, who represents that district.

The Denver pause on data centers won’t impact those that already exist or are under construction. If the council imposes new regulations, they could impact expansions, like one that CoreSite has said it plans to build on its campus.

The working group will include three council members, two representatives from utility companies, two union representatives, one industry representative, three people from advocacy groups, seven community members, one subject matter expert and seven ex officio members from Denver city departments. The council hasn’t yet announced the names of the people who will be in the group.

The group will study issues including zoning, energy use, cooling systems and renewable energy, according to a presentation.

The ordinance sponsors, council members Paul Kashmann, Darrell Watson and Gonzales-Gutierrez, are in the process of hiring a facilitator for that group’s work.

During council comments, Kashmann responded to a resident who said they opposed the moratorium because it could discourage businesses from coming to Denver.

“The message that should get out is: Regardless of the industry, Denver is definitely no longer open for unregulated businesses that — should that lack of regulation continue — present a real and significant threat to the health and welfare of the community,” Kashmann said.

The vote was unanimous, even though some council members criticized the idea of creating a task force rather than taking immediate action.

“This whole task force thing (that) I’ve seen over and over again with the administration hasn’t ever worked out,” Councilwoman Flor Alvidrez said. “So I have concerns, but the least we can do is make sure that no more are being built right now.”

Jefferson County’s moratorium does not apply to land in the county that is already zoned for data centers. Commissioner Rachel Zenzinger said in a statement Tuesday that the county has to evaluate how data centers will interact with the environment, the county’s water supply and “our community’s overall health.”

Commissioner Andy Kerr said the rules Jefferson County will hammer out over the next 10 months will look at data centers’ impacts on land, water and energy use.

In April, Weld County also updated its land-use code to address future data centers. The county commissioners passed rules that require the industry to show it has secured an adequate power and water supply before building a data center.

Tell us what you think about this story

OAKLAND, Calif. (AP) — A federal court on Monday dismissed  filed against OpenAI and its top executives by Elon Musk, who accused them of betraying a shared vision for it to remain a nonprofit dedicated to guiding artificial intelligence’s development for the good of humanity.

The nine-person jury found Musk waited too long to file his lawsuit and missed a statutory deadline. After a three-week trial, the jury deliberated less than two hours.

Musk, the world’s richest man, was a co-founder of OpenAI, which launched in 2015 and went on to create ChatGPT. After investing $38 million in its first years,  accused OpenAI  and his top deputy of shifting into a moneymaking mode behind his back.

The jury served in an advisory role, but Judge Yvonne Gonzalez Rogers accepted the verdict Monday as the court����ap own and dismissed Musk’s claims.

Musk posted on his social media platform X that he would file an appeal. He said the judge and jury never weighed in on the merits of the case, just “a calendar technicality.”

“There is no question to anyone following the case in detail that Altman & Brockman did in fact enrich themselves by stealing a charity. The only question is WHEN they did it!” he wrote.

Musk’s lawyer, Steven Molo, said Musk’s feud with OpenAI was far from resolved. He compared Monday’s verdict to moments in U.S. history like the Siege of Charleston and the Battle of Bunker Hill which were “major losses for Americans, but who won the war?”

OpenAI argued it came down to a business rivalry

The trial in Oakland, California shed light on the bitter falling-out between the two Silicon Valley titans and the beginnings of OpenAI, now  valued at $852 billion and moving toward potentially one of the largest initial public offerings in history.

Altman and OpenAI claimed there was never a promise to keep OpenAI a nonprofit forever. In fact, they argued, Musk knew this and filed his lawsuit because he couldn’t have unilateral control over the fast-growing AI developer.

OpenAI argued the lawsuit aimed to undercut the company’s rapid growth and bolster Musk’s xAI, which he launched in 2023 as a competitor.

Outside court Monday, OpenAI lawyer William Savitt told reporters that jurors determined the lawsuit was an “after-the-fact contrivance” that amounted to Musk trying to sabotage a competitor and ”to overcome a long history of very bad predictions about what OpenAI has been and will become.”

Microsoft, an OpenAI investor and a co-defendant in Musk’s lawsuit, said it welcomed the decision and remains “committed to our work with OpenAI to advance and scale AI for people and organizations around the world.”

Musk was seeking damages to be paid to the altruistic efforts of OpenAI’s charitable arm as well as Altman’s ouster from OpenAI’s board. Musk’s decision to stop funding the company contributed to the rift between the former allies. Musk says he was responding to deceptive conduct that OpenAI’s board picked up on when it  as CEO in 2023 before he  days later.

The trial saw testimony from Musk, Altman and his top lieutenant Greg Brockman, along with Microsoft CEO Satya Nadella and a slew of others in the tech titans’ orbit.

Musk told jurors on his first of three  that, fundamentally, “I think they’re going to try to make this lawsuit … very complicated, but it����ap actually very simple,” Musk said. “Which is that it����ap not OK to steal a charity.”

Musk’s lawsuit claimed that, in addition to “breach of charitable trust,” Altman and Brockman unjustly enriched themselves from the windfall as the ChatGPT maker soared in valuation. Brockman revealed during the trial that his stake in OpenAI is worth about $30 billion.

The trial detailed a clash between tech titans

Altman and Musk both vied to be OpenAI’s CEO in its early years. In his testimony, Altman said he had concerns about Musk’s attempts to gain more control over OpenAI, which was aiming to safely build a better-than-human form of AI called .

“Part of the reason we started OpenAI is we didn’t think AGI could be under the control of any one person, no matter how good their intents are,” Altman said.

The trial also shed light on Altman’s removal from the OpenAI board in 2023, before he  a few days later. Several witnesses including two ex-board members, Helen Toner and Tasha McCauley, said there were concerns about Altman’s truthfulness.

Near the end of his testimony, Altman said that before things turned sour, he had thought very highly of Musk.

“I felt like he had abandoned us, not come through on his promises, put the company in a very difficult place, jeopardized the mission, didn’t really care about the things I thought he cared about,” Altman said. “It����ap been an extremely painful thing for me … to have someone that I respected so much not acknowledge that and continue to publicly attack us.”