Utah has quietly become the first US state to let an AI chatbot renew prescriptions without a doctor, according to the Associated Press. The programme, run by a company called Doctronic, launched in January and has set off a fierce medical debate.

Residents can skip the doctor’s office and refill prescriptions online through the chatbot. It asks about their medication and history, checks a national pharmacy database, and either renews the script or escalates to a human doctor.

The launch was possible only through a “regulatory sandbox” that lets Utah officials waive laws for promising AI. State and federal rules otherwise restrict prescribing to licensed medical professionals.

“We have crossed a threshold in terms of giving something that is not human a medical license, whether or not we want to call it that,” the University of Pennsylvania’s Dr Eric Bressman told the AP. He and others say they are not opposed to AI prescribing, but want it held to standards as rigorous as those for human doctors.

The board that got left out

Utah’s medical licensing board says it only learned of the programme when the January launch made the news. In an April letter, 11 members called for the pilot to be halted, citing the risks of auto-renewing drugs with side effects or interactions.

“We were essentially told: ‘Yes this is going on. And no, you don’t have a say in it’,” said Dr Alan Smith, a family physician who chairs the board but spoke for himself.

The state declined to suspend it, noting human doctors still review every refill in this first phase.

The programme is currently overseen by a five-member board of AI specialists, none of them doctors. Doctronic expects to move to fully automated refills soon.

Smith warns the risks are real, pointing out that Doctronic’s roughly 190 refillable medications include blood thinners, which turn dangerous if a patient develops internal bleeding. The American Medical Association has echoed the concern that “prescription renewals aren’t routine checkboxes”.

A regulatory vacuum by design

The case exposes a jurisdictional tangle, since medical technology is regulated federally while medical professionals are overseen by states. Doctronic frames its AI as part of state-regulated medical practice, though some experts argue it has crossed into FDA territory.

The company would not say whether it has sought FDA permission. The agency told the AP it has authorised no AI chatbots but wants to encourage innovation, a hands-off posture that fits a broader loosening of oversight on AI health tools.

Critics see history rhyming, with Bressman comparing the moment to the haphazard medicine of the early 20th century, before boards and benchmarks existed. The template for licensing AI medical services in other states comes from the Cicero Institute, a pro-AI think tank founded by Palantir co-founder Joe Lonsdale.

The stakes are not abstract, as safety researchers have warned that medical chatbots can sound authoritative while dispensing dangerous advice. Others caution that removing humans from care can undermine the very outcomes it promises.

Rivals are scrambling to map those failure modes too. Meta went as far as posing as teenagers to test how competing chatbots handle sensitive topics.

Doctronic plans peer-reviewed studies later this year, though its only published paper so far was written by its own scientists and not independently reviewed. As one Utah law professor put it, companies risk letting the technology race beyond the evidence, and betraying public trust in the process.

A federal judge has refused to overturn a jury’s finding that Elon Musk defrauded Twitter investors during his $44bn takeover of the platform in 2022. US District Judge Charles Breyer denied Musk’s motion to set aside the verdict in most respects on Monday.

A San Francisco jury ruled in March that two of Musk’s May 2022 tweets about the deal and Twitter’s spam bot numbers were materially false or misleading. Investors say the resulting losses could support damages of up to $2.6bn.

“Buyer’s remorse is not an exception to the securities laws,” Breyer wrote, adding that the laws are “in their essence, about trust”. The judge found substantial evidence that Musk’s 13 May tweet, claiming the deal was on hold pending bot data, was literally untrue.

Breyer cited testimony from one of Musk’s own bankers, who said the tweet surprised her and that Musk never actually put the deal on hold. A jury could infer Musk had a motive to escape the deal and used bots as a pretext, the judge wrote.

He did hand Musk one narrow win, agreeing there was too little evidence that a separate 17 May tweet caused investors a market loss. Musk’s lawyers did not immediately respond to requests for comment.

The bot pretext, four years on

The case traces back to Musk’s chaotic pursuit of Twitter, when he agreed to buy the company, then tried to walk away citing spam accounts. Twitter sued to force the deal through, and Musk ultimately closed at $54.20 a share before renaming the platform X.

Investors sued in October 2022, arguing Musk deliberately talked the stock down to renegotiate or exit. The jury agreed he misled the market, though it rejected the broader claim that he ran a deliberate scheme.

Breyer also swatted down Musk’s more colourful arguments, including a claim that jurors mocked him by writing “$4.20” in blue ink on the verdict form. The number references cannabis, the judge noted, and the jury had actually cleared Musk on two claims.

Another legal front for a busy defendant

The ruling adds to a crowded docket for Musk, who recently settled a separate SEC case over his late disclosure of an initial Twitter stake for $1.5m. His “funding secured” Tesla saga first drew SEC fraud charges back in 2018.

He is also fighting Sam Altman in a high-stakes trial over OpenAI, all while steering the newly public SpaceX. The tweets that built his mythology keep generating legal bills.

Prejudgment interest, which Breyer also granted, could push the final figure higher still. For a man now worth more than a trillion dollars, the sum is survivable, but the finding that he defrauded investors is harder to shrug off.

from the here-to-build-an-encyclopedia dept

In Larry Sanger’s recent failed attempt to start a “WikiProject Intellectual Diversity”, he tried to recruit his followers to help him change Wikipedia’s rules around representation of viewpoints, religions, parties, and nationalities (a version of his earlier “Nine Theses”). The draft WikiProject was not itself a bannable offense, but his approach broke rules that were designed to foster fair discussions. Wikipedia’s rules really already support creation of balanced and robust articles about controversial topics – it just takes a huge amount of careful research, patience, and cooperation, and there’s no shortcut for that work.

In the first several months of Wikipedia, Sanger’s seriousness about its potential encouraged me to take up the challenge of helping write an encyclopedia that represents the sum of human knowledge. 25 years later, I remain an active editor dedicated to the Wikimedia movement for free and open knowledge, which is basically a fun and oddly serious hobby.

I edit a lot of moderately controversial articles that have glaring gaps in core principles of verifiability and neutral point of view. Many of Wikipedia’s most popular articles, like about politics and philosophy, are very informative and comprehensive, but second-tier articles don’t consistently get robust attention from editors. For example, I’ve recently repaired bias and disinformation in articles about AI regulation, LGBTQ rights in Nigeria, politicians in the Balkans, wealthy businessmen outside the US, influential religious organizations, and people accused of sexual harassment. I routinely fix articles that downplay negative information or present a controversial topic in a flattering way, in the style of Jeffrey Epstein’s ineffective project to get consultants to sanitize his article.

The good thing is that Wikipedia’s established rules already provide robust strategies to improve verifiability and balance in articles. Its principles expect editors to be cooperative and willing to cite a reliable source for nearly every sentence. You have to be up for changing your mind when somebody finds multiple reliable sources that disprove something you assumed, or at least up for slinking away to another article. To help counter bias and conflicts of interest, I apply elaborately layered guidance for evaluating and weighing sources – often citing academic journal articles and books, but not always, because the guidance recognizes that reliability is contextual. The “due weight” policy, part of the neutral point of view policy, pushes editors to search for more and better sources when something gets disputed, which results in a stronger article. I’ve learned that the best way to resolve a content dispute is to cite the best sources, reference the most relevant rules, present evidence calmly, and escalate one step at a time through the dispute resolution forums. Dispute resolution typically uses Wikipedia’s informal decision-making process, which reflects that Wikipedia is a decentralized asynchronous volunteer project, not an adjudicatory body. Wikipedia’s processes already work pretty well, they just take a lot of skill and patience, because collaboration is hard work.

Sanger was banned for off-Wikipedia canvassing and for not being on Wikipedia to build an encyclopedia, but to be clear, trying to start WikiProject Intellectual Diversity was not in itself a bannable offense. Canvassing is against the rules specifically to protect public and open processes that support the development of balanced articles. The canvassing guidelines discourage editors from trying to rig decision-making processes by selectively inviting participants who will take their side. The rules favor public discussions on Wikipedia so that all editors have an equal opportunity to participate. And since all Wikipedia edits are publicly tracked, editors can analyze each other’s contributions to detect biases and conflicts of interest. External invitations both selectively invite participation and prevent editors from exercising oversight. Volunteer administrators routinely block or even ban editors for inappropriate canvassing because this behavior compromises efforts to build a balanced encyclopedia.

Sanger’s recent advocacy reminds me of the pattern that researcher and Wikipedia editor Molly White described in January 2025: “right-wing voices attacking Wikipedia as part of an intensifying campaign against free and open access information.” In October, the Washington Post described Sanger as “fueling the right’s campaign” against Wikipedia. Among other incidents last year, House Republicans demanded disclosure of editor information over coverage of the Israel-Palestine conflict. Sanger’s call to prohibit anonymity for the most dedicated volunteer administrators, one of his Nine Theses, is another one of his takes that would undermine intellectual freedom in the project, in line with the leaked Heritage Foundation plan to dox editors.

My work to counter gaps, bias, and spam in Wikipedia articles gives me proof every day that the project is imperfect. Every active editor has critiques of Wikipedia, the Wikimedia Foundation, and the Wikimedia movement, and we debate issues and improvements at length. Wikipedia would benefit from additional contributors from any viewpoint or background who want to help build an encyclopedia. But improving Wikipedia requires intellectual honesty, cooperation, and willingness to apply established principles and rules even while critiquing them, not bad-faith publicity stunts.

Filed Under: canvassing, diversity, encyclopedia, larry sanger, wikipedia

Companies: wikimedia foundation, wikipedia

Whether you subscribe to the 2TB iCloud+ tier individually or pay for Apple One Premier, you’re getting the new Apple Intelligence in Apple Home features announced during WWDC 2026.

Apple didn’t break out exactly what customers might have to pay in order to access its most advanced AI features. While there aren’t any separate AI subscriptions or token purchase programs, users will need to spend more cash for the most access.

In the macOS Golden Gate beta release notes, Apple has confirmed that the Apple Home AI features will require the 2TB iCloud+ plan. On its own, that is a $10 a month plan, or is included with the $37.95 Apple One Premier subscription tier.

Either way, customers already paying for these products will gain more Apple Home features. The 2TB iCloud+ plan was already required to have unlimited cameras for HomeKit Secure Video.

While Apple isn’t being clear about which features are being lumped in here, it seems it is just referring to the HomeKit Secure Video AI analysis feature. It analyzes footage locally for people, objects, and events to piece together what happens in a recorded clip.

In the Apple Home app, those events can be stitched together into a series of clips, or shown as priority events. Either way, it is meant to serve as an easier way to review and search video recordings.

The notification grouping and 4K HomeKit Secure Video support don’t appear to be tied to the subscription, since they’re not relying on Apple’s AI image models. The lower 200GB and 50GB tiers are limited to 5 cameras and 2 cameras respectively.

Apple is making it clear: if you want full access to its AI and cloud features, you need to pay for its more premium services. It isn’t clear if Apple plans to offer separate AI subscriptions in the future.

A phishing campaign is impersonating more than 30 well-known brands, including Adobe, Netflix, Coca-Cola, and OpenAI, in fake job interviews to steal Google account credentials from marketing professionals.

The operation is abusing the legitimate cloud-based PeopleForce human resources platform and a domain associated with the Salesforce Marketing Cloud service before redirecting the recipient to a malicious landing page.

To further instill trust and increase the chances of success, the threat actor is using the names and pictures of real recruiters at impersonated companies.

Will Thomas, senior advisor at cybersecurity intelligence and threat hunting company Team Cymru, analyzed the campaign and discovered that the phishing email pretends to be from “a recruiter looking to hire people for marketing roles.”

The researcher uncovered that the threat actor is using at least 34 domains impersonating high-value companies in the following sectors:

• Airlines and travel: American Airlines, Booking.com, Delta Air Lines, United Airlines
• Food and beverage: Coca-Cola, PepsiCo, Red Bull
• Apparel and luxury goods: Adidas, Louis Vuitton, Sephora, Levis
• Staffing, consulting, and tech: Adobe, Aquent, ManpowerGroup, McKinsey & Company, OpenAI
• Hospitality and marketing: Marriott, Omnicom Group
• Entertainment and sports: FIFA, Netflix

Thomas found that the campaign relies on nested redirects, a technique that routes visitors through multiple legitimate services before reaching the malicious landing page.

The researcher noted that while the phishing emails appear to originate from PeopleForce, the underlying links resolve to the exct[.]net domain, which is operated by Salesforce following its acquisition of the ExactTarget marketing automation platform, now rebranded as Salesforce Marketing Cloud.

ExactTarget redirects to the Wise Agent (wiseagent[.]com) cloud-based real estate Customer Relationship Management (CRM) software for agents, teams, and brokers, which forwards to the phishing landing page.

BleepingComputer has found that the operation has been running for at least five months and initially used Outlook email addresses with the name of the impersonated company.

One phishing email, posing as a message from Adidas recruiter Paulina Manzo, asked the recipient to schedule a conversation about a potential role at the company.

When clicking on the link to the calendar, the recipient was redirected to the threat actor’s landing page adidas-hiring[.]com

To continue the process for scheduling a meeting with the recruiter, the potential victim is asked to sign into their Google account.

Clicking on the “Continue with Google” button triggers a fake Google sign-in popup rendered inside the phishing page to impersonate Google authentication.

Although the pop-up may appear as a legitimate browser window, it is just HTML and CSS code rendered inside the phishing page, a technique known as browser-in-the-browser (BitB).

By using modern web development tools, the attacker can imitate all the elements of a legitimate authentication pop-up page.

While it is unclear how the threat actor gained access to the legitimate platforms, abusing them does not imply a compromise of the services.

One possible avenue is creating a genuine account specifically for the campaign, or using compromised logins, which allows configuring the redirect chain and the landing page.

A list of the domains discovered in this phishing campaign is available in Will Thomas’ analysis on GitHub.

Security teams log 54% of successful attacks and alert on just 14%. The rest move through your environment unseen.

The Picus whitepaper shows how breach and attack simulation tests your SIEM and EDR rules so threats stop slipping by detection.

Get the whitepaper

Businesses looking to modernize their payment systems may want to learn what a crypto payment processor is, how it works, and why it matters.

Cryptocurrency and the crypto industry at large often appear most prominently in investing circles, but recent advancements in blockchain, the technology that supports crypto, have expanded crypto’s usefulness. Today, businesses across industries can make practical use of crypto thanks to the advent of crypto payment processors like 0xProcessing, tools that allow them to accept crypto payments without holding crypto themselves.

Crypto and related technologies remain something of a mystery for many companies, however, so to better determine whether adopting a crypto payment processor is the right move for them, businesses should learn more about what these tools are, how they work, and why they may prove useful.

What Is a Crypto Payment Processor?

Put simply, a crypto payment processor is a tool that allows businesses to accept various cryptocurrencies as payment. When a customer selects crypto at checkout, the processor generates a unique wallet address or QR code for that transaction. The customer sends their payment to that address for verification, and upon successful verification, the processor either converts the funds to the business’s preferred fiat currency or deposits them directly into the business’s crypto wallet.

Note that this approach means businesses do not have to accept crypto if they do not want to; if a company would prefer not to deal with the volatility of managing Bitcoin, for instance, they can simply convert it to USD at the counter, minimizing the risk of the payment losing value.

Transactions made using crypto can take anywhere from a few seconds to a few minutes, depending on the cryptocurrency used, so businesses should prepare accordingly.

Why Businesses Might Use a Crypto Payment Processor

Aside from expanding payment options, there are several potential benefits to using a crypto payment processor. For businesses that operate online, accepting crypto may give them access to global customers, since most cryptocurrencies can be used from almost anywhere.

According to Deloitte, “Using crypto as a form of payment could reduce transaction fees and possibly eliminate the cost of float and the need to wait multiple days for cash settlement.”

This could be particularly true for cross-border payments, which have historically been more costly and take longer to process than domestic payments.

In more urban regions where competition between businesses tends to be fiercer, expanding the currencies they accept can make a given business appear more innovative, making it more appealing to tech-savvy consumers and other customers who may prefer to pay in alternative currencies.

Picking the Right Processor

Although crypto payment processors generally have benefits, businesses should be mindful of the specific processor they adopt, as they vary in overall quality and security.

On the whole, reliable crypto payment processors employ security features such as blockchain transparency, fraud prevention, wallet security, and compliance and KYC/AML measures. No security feature is foolproof, but these protocols may help businesses keep both their data and their customers’ data safe from fraud. This priority will likely become increasingly important as more purchases are made digitally.

Before adopting a crypto payment processor, businesses should thoroughly assess whether it would benefit them. If they serve an audience that is not interested in paying with crypto, chances are that tools to help them do so would have limited use.

Though cryptocurrency is by no means ubiquitous as a regular form of payment, its increasing popularity could prompt businesses that previously disregarded it to consider its implications for their future. For businesses that see crypto as a practical asset in the years to come, adopting a crypto payment processor could be a solid first step in preparing their operations for that future.

FAQ

Q: What is a crypto payment processor?
A: A crypto payment processor enables businesses to accept cryptocurrency payments while simplifying transaction management and settlement.

Q: Is it safe to accept cryptocurrencies as payment?
A: Generally yes, as reputable processors use blockchain verification, security controls, and compliance measures to help protect transactions.

Q: Can businesses receive fiat instead of cryptocurrency?
A: Indirectly, yes. Many crypto payment processors automatically convert crypto payments into traditional currencies.

Q: Which industries benefit most from crypto payments?
A: E-commerce, SaaS, gaming, travel, and digital service providers often see the greatest benefits from accepting cryptocurrency payment

Known for its cloud infrastructure that allows developers to deploy agents without managing servers, Vercel has quietly become one of the most central companies in AI software. The company currently sees 6 million deployments a day, half of them triggered by coding agents, and more than 1 trillion tokens flow through the company’s AI gateway daily.

After the company’s ShipNYC conference last week, we sat down with Vercel CEO Guillermo Rauch for his take on this moment in AI, and how platform companies like Vercel end up competing with major labs. Here’s a lightly edited transcript.

It feels like there’s a different energy in the community this year, fewer pilot programs and more focus on how to make things work well in practice. I’m sure you’ve seen that a lot with clients, but I’m curious what that journey has looked like within Vercel.

Last year was about prototyping. The sky’s the limit, unleash the agents, everyone can build, and so on. We did that, and we learned a lot because we had hundreds of agents organically developed and deployed within the company, and then you started getting into the realities of agents in production, and some of the challenges.

The biggest lesson for me was the home-run use cases, the two killer apps of agents. One is the coding agent, of course. That’s driving a lot of the token utilization in the world, but when you produce so much software, you need somewhere to put it. The second killer app of agents is the internal agent that helps you run the company. The challenge there is, how do you securely access data? How do you audit what the agent is doing? How do you get a trail of all of the tool calls and access controls that the agent had to incur in order to get a job done?

To solve that, we came up with this framework called Eve, where you can lay out an agents’ instructions and skills in natural language. And another tool is Vercel Sandbox, where you put the agent in a little cage. It can have the freedom still to express its intelligence, but then you can apply policy on what data it can access and what data can leave the sandbox.

What sort of problems does that help you avoid?

For [the] sandbox, the biggest advantage is data control. A real risk of AI that I always think about is, when you get a coding IDE like Devin or Cursor, if you’re in the wrong setting, they may train on your entire codebase. I remember talking to the president of Airbus about this. You have decades of wealth of very specific C++ code for aerospace engineering. Someone comes in and installs the wrong developer tool and boom, all the code goes out to the cloud for training.

I’m curious to hear more about that second killer use case. We all know about coding agents, but what does an internal corporate agent look like in practice?

So, there’s a sales rep sitting out there [in Vercel’s office]. She works on install base. Her job is to grow existing accounts. The bottleneck for people like her has not been her creativity, intelligence, ability to build relationships, it’s been data. “I don’t understand what accounts are growing faster. Give me the five accounts that have added the most seats in the last two weeks, so that I can prioritize my work.” She couldn’t ask that question in the past. She needed to wait until a Q1 project for a new sales dashboard completed. 

We were in that bottleneck for years at Vercel, and it was really frustrating because on the R&D side, we’re the fastest-moving company in the world. But on the sales engine, the Salesforce engineering [side], I was so incompetent. I had never opened Salesforce in my life when I started.

Now I feel like I can actually have impact across the entire company, because Eve can be used for our customer-facing agents and can be used to improve productivity. Same technology, it’s just APIs. Agents are forcing companies to open up, and that will have dramatic long-term implications. So many of these SaaS giants build their entire kingdoms on trapping your data, and that’s incompatible with agents.

How do you see client relationships with the big AI labs changing?

Last year there were a lot of people picking one lab partner — saying they would build everything on OpenAI or Anthropic. Now they’re saying, I understand how this all works — model, harness, data platform, sandbox, gateway — every piece is plug and play. You can use OpenAI, you can use Anthropic, or you can use Gemini. We’re seeing a lot of growth of Gemini, even though it’s not on the news as much, because people are optimizing for production now. The reality is, when you’re optimizing for production, you start looking at a price/performance, and Gemini models have awesome price/performance characteristics. You also bring in open models, so DeepSeek and GLM-5.2 are taking off. The data doesn’t lie.

There are places where you’re in direct competition with the labs too, right? Just the other week, OpenAI released a new set of tools that publish directly to the web without having to leave the OpenAI enclave.

It’s a natural next step for them to host little websites. And it’s a great opening for us, because now people will think of ChatGPT as a tool for making websites. And then if they keep asking the model questions about web hosting, the model recommends us. But you’re right, as the models or platforms add more capabilities, they come in direct competition with the infrastructure platforms that already exist.

I really think at this point we’re deciding on whether the model and the agent are going to be coupled.

Do you get all your intelligence from one place? Or do you get a module or a library or a building block from one provider, and then you build on top of it. That’s more like software engineering has always been, and that’s really what we’re bringing to market. We’re going to be the AWS of this generation, so obviously we’re fighting for a world of open protocols.

​Vietnamese authorities have arrested and are prosecuting seven suspects believed to have run HiAnime, the largest anime piracy streaming service before its shutdown in June.

HiAnime provided access to a massive library of English-subbed and dubbed anime without subscription fees, attracting several hundred million visitors each month and temporarily surpassing legal streaming platforms like Disney+ and Crunchyroll in web traffic between late 2024 and 2025.

It was launched on the Zoro.to domain, rebranded to Aniwatch (and switched to Aniwatch.to) in July 2023, and again in March 2024 as HiAnime/H!Anime (using the HiAnime.to domain).

After becoming massively popular, HiAnime was also placed on the European Commission’s Counterfeit and Piracy Watch List and the United States Trade Representative’s (USTR) Notorious Markets list.

The seven defendants have been charged with infringing copyright and related rights and with money laundering, with four of them detained and the other three placed under house arrest.

They have been accused of creating more than 100 websites to upload over 26,000 pirated anime films, generating approximately $12.85 million in illegal advertising revenue between 2020 and April 2026.

​The Alliance for Creativity and Entertainment (ACE), a coalition of over 50 media and entertainment companies, including the world’s largest film studios and television networks, focused on shuttering illegal streaming services, confirmed the law enforcement action on Thursday and thanked U.S. authorities for their support throughout a multi-year investigation that led to the suspects’ arrests.

​”ACE applauds the actions of Vietnam’s Ministry of Public Security (MPS), in particular C03, the Economic Crimes Investigation Department, and A05, the Department of Cybersecurity and High-Tech Crime Prevention, in arresting and prosecuting seven operators believed to be behind Hianime and related piracy services,” said the Alliance for Creativity and Entertainment on Thursday.

“ACE would also like to thank its partners, Homeland Security Investigations and the U.S. Department of Justice, for their continued support in this multi-year investigation and action. ACE looks forward to continuing to support the MPS and its relevant agencies, and to working even more closely with them on future actions against piracy services.”

Earlier this year, in March, ACE also announced the shutdown of AnimePlay, another major anime streaming platform that hosted more than 60 terabytes of anime TV shows and movies and had over 5 million registered users.

The anti-piracy coalition dismantled AnimePlay by taking all infrastructure offline, including its hosting servers and web domains.

Security teams log 54% of successful attacks and alert on just 14%. The rest move through your environment unseen.

The Picus whitepaper shows how breach and attack simulation tests your SIEM and EDR rules so threats stop slipping by detection.

Get the whitepaper