On Thursday, Anthropic unveiled specialized AI models designed for US national security customers. The company released "Claude Gov" models that were built in response to direct feedback from government clients to handle operations such as strategic planning, intelligence analysis, and operational support. The custom models reportedly already serve US national security agencies, with access restricted to those working in classified environments.

The Claude Gov models differ from Anthropic's consumer and enterprise offerings, also called Claude, in several ways. They reportedly handle classified material, "refuse less" when engaging with classified information, and are customized to handle intelligence and defense documents. The models also feature what Anthropic calls "enhanced proficiency" in languages and dialects critical to national security operations.

Anthropic says the new models underwent the same "safety testing" as all Claude models. The company has been pursuing government contracts as it seeks reliable revenue sources, partnering with Palantir and Amazon Web Services in November to sell AI tools to defense customers.

Read full article

Comments

© Anthropic

On Thursday, Anthropic CEO Dario Amodei argued against a proposed 10-year moratorium on state AI regulation in a New York Times opinion piece, calling the measure shortsighted and overbroad as Congress considers including it in President Trump's tax policy bill. Anthropic makes Claude, an AI assistant similar to ChatGPT.

Amodei warned that AI is advancing too fast for such a long freeze, predicting these systems "could change the world, fundamentally, within two years; in 10 years, all bets are off."

As we covered in May, the moratorium would prevent states from regulating AI for a decade. A bipartisan group of state attorneys general has opposed the measure, which would preempt AI laws and regulations recently passed in dozens of states.

Read full article

Comments

© Bloomberg via Getty Images

Over the years, scholars of the Dead Sea Scrolls have analyzed the ancient parchments with various methods: for example, X-rays, multispectral imaging, "virtual unfolding," and paleography, i.e., studying elements in their writing styles. The scrolls are believed to date back to between the third century BCE and the first century CE, but those dates rely largely on paleography, since only a handful of the scrolls have calendar dates written on them.

However, the traditional paleographic method is inherently subjective and based on a given scholar's experience. A team of scientists has combined radiocarbon dating from 24 scroll samples and machine-learning-based handwriting analysis to create their own AI program—dubbed Enoch. The objective was to achieve more accurate date estimates, according to a new paper published in the journal PLoS ONE. Among the findings: Many of the scrolls are older than previously thought.

As reported earlier, these ancient Hebrew texts—roughly 900 full and partial scrolls in all, stored in clay jars—were first discovered scattered in various caves near what was once the settlement of Qumran, just north of the Dead Sea, by Bedouin shepherds in 1946–1947. (Apparently, a shepherd threw a rock while searching for a lost member of his flock and accidentally shattered one of the clay jars, leading to the discovery.) Qumran was destroyed by the Romans, circa 73 CE, and historians believe the scrolls were hidden in the caves by a sect called the Essenes to protect them from being destroyed. The natural limestone and conditions within the caves helped preserve the scrolls for millennia.

Read full article

Comments

© Michael Kappeler/AFP/Getty Images

On Sunday, independent AI researcher Simon Willison published a detailed analysis of Anthropic's newly released system prompts for Claude 4's Opus 4 and Sonnet 4 models, offering insights into how Anthropic controls the models' "behavior" through their outputs. Willison examined both the published prompts and leaked internal tool instructions to reveal what he calls "a sort of unofficial manual for how best to use these tools."

To understand what Willison is talking about, we'll need to explain what system prompts are. Large language models (LLMs) like the AI models that run Claude and ChatGPT process an input called a "prompt" and return an output that is the most likely continuation of that prompt. System prompts are instructions that AI companies feed to the models before each conversation to establish how they should respond.

Unlike the messages users see from the chatbot, system prompts typically remain hidden from the user and tell the model its identity, behavioral guidelines, and specific rules to follow. Each time a user sends a message, the AI model receives the full conversation history along with the system prompt, allowing it to maintain context while following its instructions.

Read full article

Comments

© AndreyPopov via Getty Images

On Thursday, Anthropic released Claude Opus 4 and Claude Sonnet 4, marking the company's return to larger model releases after primarily focusing on mid-range Sonnet variants since June of last year. The new models represent what the company calls its most capable coding models yet, with Opus 4 designed for complex, long-running tasks that can operate autonomously for hours.

Alex Albert, Anthropic's head of Claude Relations, told Ars Technica that the company chose to revive the Opus line because of growing demand for agentic AI applications. "Across all the companies out there that are building things, there's a really large wave of these agentic applications springing up, and a very high demand and premium being placed on intelligence," Albert said. "I think Opus is going to fit that groove perfectly."

Before we go further, a brief refresher on Claude's three AI model "size" names (introduced in March 2024) is probably warranted. Haiku, Sonnet, and Opus offer a tradeoff between price (in the API), speed, and capability.

Read full article

Comments

© Anthropic

For a short period of time on Friday, Darth Vader could drop F-bombs in the video game Fortnite as part of a voice AI implementation gone wrong, reports GameSpot. Epic Games rapidly deployed a hotfix after players encountered the Sith Lord responding to their comments with profanity and strong language.

In Fortnite, the AI-voiced Vader appears as both a boss in battle royale mode and an interactive character. The official Star Wars website encourages players to "ask him all your pressing questions about the Force, the Galactic Empire… or you know, a good strat for the last Storm circle," adding that "the Sith Lord has opinions."

The F-bomb incident involved a Twitch streamer named Loserfruit, who triggered the forceful response when discussing food with the virtual Vader. The Dark Lord of the Sith responded by repeating her words "freaking" and "fucking" before adding, "Such vulgarity does not become you, Padme." The exchange spread virally across social media platforms on Friday.

Read full article

Comments

© Disney / Starwars.com

---

Google's AlphaEvolve is the epitome of a best-practice AI agent orchestration. It offers a lesson in production-grade agent engineering. Discover its architecture & essential takeaways for your enterprise AI strategy.Read More

On Wednesday, OpenAI announced that ChatGPT users now have access to GPT-4.1, an AI language model previously available only through the company's API since its launch one month ago. The update brings what OpenAI describes as improved coding and web development capabilities to paid ChatGPT subscribers, with wider enterprise rollout planned in the coming weeks.

Adding GPT-4.1 and 4.1 mini to ChatGPT adds to an already complex model selection that includes GPT-4o, various specialized GPT-4o versions, o1-pro, o3-mini, and o3-mini-high models. There are technically nine AI models available for ChatGPT Pro subscribers. Wharton professor Ethan Mollick recently publicly lampooned the awkward situation on social media.

As of May 14, 2025, ChatGPT Pro users have access to eight main AI models, plus Deep Research. Credit: Benj Edwards

Deciding which AI model to use can be daunting for AI novices. Reddit users and OpenAI forum members alike commonly voice confusion about the available options. "I do not understand the reason behind having multiple models available for use," wrote one Reddit user in March. "Why would anyone use anything but the best one?" Another Redditor said they were "a bit lost" with the many ChatGPT models available after switching back from using Anthropic Claude.

Read full article

Comments

© Getty Images

Using AI can be a double-edged sword, according to new research from Duke University. While generative AI tools may boost productivity for some, they might also secretly damage your professional reputation.

On Thursday, the Proceedings of the National Academy of Sciences (PNAS) published a study showing that employees who use AI tools like ChatGPT, Claude, and Gemini at work face negative judgments about their competence and motivation from colleagues and managers.

"Our findings reveal a dilemma for people considering adopting AI tools: Although AI can enhance productivity, its use carries social costs," write researchers Jessica A. Reif, Richard P. Larrick, and Jack B. Soll of Duke's Fuqua School of Business.

Read full article

Comments

© demaerre via Getty Images

On Wednesday, OpenAI announced that Instacart CEO Fidji Simo will join the maker of ChatGPT as "CEO of Applications" later this year, according to a company blog post. Simo, who has served on the company's board since March 2024, will oversee business and operational teams while continuing to report directly to Altman in the newly created role. Altman will remain the primary CEO of OpenAI.

According to Reuters, Simo spent a decade at Meta, including a stint serving as the head of Facebook from 2019 to 2021. She also currently sits on the board of e-commerce services site Shopify.

The announcement came earlier than planned due to what Altman described as "a leak" that "accelerated our timeline." At OpenAI, Simo will manage what Altman called "traditional company functions" as the organization enters its "next phase of growth." The applications category at OpenAI includes products like ChatGPT, the popular AI assistant.

Read full article

Comments

© Joel Saget via Getty Images

On Wednesday, the Trump administration announced plans to rescind and replace a Biden-era rule regulating the export of high-end AI accelerator chips worldwide, Bloomberg and Reuters reported.

A Department of Commerce spokeswoman told Reuters that officials found the previous framework "overly complex, overly bureaucratic, and would stymie American innovation" and pledged to create "a much simpler rule that unleashes American innovation and ensures American AI dominance."

The Biden administration issued the Framework for Artificial Intelligence Diffusion in January during its final week in office. The regulation represented the last salvo of a four-year effort to control global access to so-called "advanced" AI chips (such as GPUs made by Nvidia), with a focus on restricting China's ability to obtain tech that could enhance its military capabilities.

Read full article

Comments

© SEAN GLADWELL via Getty Images

On Monday, ChatGPT-maker OpenAI announced it will remain under the control of its founding nonprofit board, scrapping its controversial plan to split off its commercial operations as a for-profit company after mounting pressure from critics.

In an official OpenAI blog post announcing the latest restructuring decision, CEO Sam Altman wrote: "We made the decision for the nonprofit to stay in control after hearing from civic leaders and having discussions with the offices of the Attorneys General of California and Delaware."

The move represents a significant shift in OpenAI's proposed restructuring. While the most recent previous version of the company's plan (which we covered in December) would have established OpenAI as a Public Benefit Corporation with the nonprofit merely holding shares and having limited influence, the revised approach keeps the nonprofit firmly in control of operations.

Read full article

Comments

© Benj Edwards / OpenAI

---

The answer to 'What customer needs requires an AI solution?' isn’t always 'Yes.' LLMs are still expensive and not always accurate.Read More

On Thursday, OpenAI announced the addition of shopping features to ChatGPT Search. The new feature allows users to search for products and purchase them through merchant websites after being redirected from the ChatGPT interface. Product placement is not sponsored, and the update affects all users, regardless of whether they've signed in to an account.

Adam Fry, ChatGPT search product lead at OpenAI, showed Ars Technica's sister site Wired how the new shopping system works during a demonstration. Users researching products like espresso machines or office chairs receive recommendations based on their stated preferences, stored memories, and product reviews from around the web.

According to Wired, the shopping experience in ChatGPT resembles Google Shopping. When users click on a product image, the interface displays multiple retailers like Amazon and Walmart on the right side of the screen, with buttons to complete purchases. OpenAI is currently experimenting with categories that include electronics, fashion, home goods, and beauty products.

Read full article

Comments

© Westend61 via Getty Images

There's a curious contradiction at the heart of today's most capable AI models that purport to "reason": They can solve routine math problems with accuracy, yet when faced with formulating deeper mathematical proofs found in competition-level challenges, they often fail.

That's the finding of eye-opening preprint research into simulated reasoning (SR) models, initially listed in March and updated in April, that mostly fell under the news radar. The research serves as an instructive case study on the mathematical limitations of SR models, despite sometimes grandiose marketing claims from AI vendors.

What sets simulated reasoning models apart from traditional large language models (LLMs) is that they have been trained to output a step-by-step "thinking" process (often called "chain-of-thought") to solve problems. Note that "simulated" in this case doesn't mean that the models do not reason at all but rather that they do not necessarily reason using the same techniques as humans. That distinction is important because human reasoning itself is difficult to define.

Read full article

Comments

© PhonlamaiPhoto via Getty Images

Ironically, our present AI age has shone a bright spotlight on the immense value of human creativity as breakthroughs in technology threaten to undermine it. As tech giants rush to build newer AI models, their web crawlers vacuum up creative content, and those same models spew floods of synthetic media, risking drowning out the human creative spark in an ocean of pablum.

Given this trajectory, AI-generated content may soon exceed the entire corpus of historical human creative works, making the preservation of the human creative ecosystem not just an ethical concern but an urgent imperative. The alternative is nothing less than a gradual homogenization of our cultural landscape, where machine learning flattens the richness of human expression into a mediocre statistical average.

A limited resource

By ingesting billions of creations, chatbots learn to talk, and image synthesizers learn to draw. Along the way, the AI companies behind them treat our shared culture like an inexhaustible resource to be strip-mined, with little thought for the consequences.

Read full article

Comments

© Kenny McCartney via Getty Images

On Monday, the State Bar of California revealed that it used AI to develop a portion of multiple-choice questions on its February 2025 bar exam, causing outrage among law school faculty and test takers. The admission comes after weeks of complaints about technical problems and irregularities during the exam administration, reports the Los Angeles Times.

The State Bar disclosed that its psychometrician (a person or organization skilled in administrating psychological tests), ACS Ventures, created 23 of the 171 scored multiple-choice questions with AI assistance. Another 48 questions came from a first-year law student exam, while Kaplan Exam Services developed the remaining 100 questions.

The State Bar defended its practices, telling the LA Times that all questions underwent review by content validation panels and subject matter experts before the exam. "The ACS questions were developed with the assistance of AI and subsequently reviewed by content validation panels and a subject matter expert in advance of the exam," wrote State Bar Executive Director Leah Wilson in a press release.

Read full article

Comments

© Getty Images

---

Based on hands-on comparisons and real-world tests, this guide unpacks what happens when you switch from OpenAI to Anthropic or Google’s Gemini and what your team needs to watch for.Read More

On Wednesday, OpenAI announced the release of two new models—o3 and o4-mini—that combine simulated reasoning capabilities with access to functions like web browsing and coding. These models mark the first time OpenAI's reasoning-focused models can use every ChatGPT tool simultaneously, including visual analysis and image generation.

OpenAI announced o3 in December, and until now, only less capable derivative models named "o3-mini" and "03-mini-high" have been available. However, the new models replace their predecessors—o1 and o3-mini.

OpenAI is rolling out access today for ChatGPT Plus, Pro, and Team users, with Enterprise and Edu customers gaining access next week. Free users can try o4-mini by selecting the "Think" option before submitting queries. OpenAI CEO Sam Altman tweeted that "we expect to release o3-pro to the pro tier in a few weeks."

Read full article

Comments

© Floriana via Getty Images

In the AI world, a vulnerability called a "prompt injection" has haunted developers since chatbots went mainstream in 2022. Despite numerous attempts to solve this fundamental vulnerability—the digital equivalent of whispering secret instructions to override a system's intended behavior—no one has found a reliable solution. Until now, perhaps.

Google DeepMind has unveiled CaMeL (CApabilities for MachinE Learning), a new approach to stopping prompt-injection attacks that abandons the failed strategy of having AI models police themselves. Instead, CaMeL treats language models as fundamentally untrusted components within a secure software framework, creating clear boundaries between user commands and potentially malicious content.

The new paper grounds CaMeL's design in established software security principles like Control Flow Integrity (CFI), Access Control, and Information Flow Control (IFC), adapting decades of security engineering wisdom to the challenges of LLMs.

Read full article

Comments

© Aman Verma via Getty Images