Daily AI Briefing

Both defenses blocked most attacks. With fewer tools, the agent got 73% of its normal tasks right. With the detector, it got only 41%. Continue reading on Towards AI »

AI generated code ships fast, but the maintenance data tells a different story. Here’s what comprehension debt actually costs. Continue reading on Towards AI »

Six documented incidents in 12 months show how AI agents went from hacking assistants to autonomous attackers. Continue reading on Towards AI »

Pro-datacentre officials argue scale of Australia’s buildout is smaller, more regulated and less damaging than in US Australian politicians and business leaders are arguing that their populace’s growing resentment of AI datacentres is an inauthentic import from the United States. Like the US, Australia is seeing community resentment grow in response to a surge in the construction of massive AI datacentres. However, pro-datacentre officials and executives argue the scale of Australia’s buildout i

Why Your AI Agent Needs a Domain Model Before It Needs a Better Prompt Continue reading on Towards AI »

Continue reading on Towards AI »

Blue Cross Blue Shield says hospital use of AI tools led to an additional $942M in healthcare spending over a two-year period.

Drums on film. | Image: Decap I don't think I'm going to hurt anyone's feelings by pointing out that Decap doesn't have the name recognition of Kendrick Lamar, Olivia Rodrigo, or Charli XCX. But his fingerprints are all over tracks from them, as well as many other artists, including Freddie Gibbs, Bad Bunny, Mac Miller, BTS, and even Ludwig Göransson's score for Oppenheimer. Nicholas Piantedosi, better known as Decap, is a record producer, but perhaps more importantly, he's the man behind Dru

Middle schoolers, likely blocked from other apps and social networks, apparently turned the Spotify comment section under an episode of NPR's Wild Card into an impromptu group chat. In a new episode of This American Life, host Ira Glass digs into how the kids landed there and why. But it mostly boils down to kids getting creative to work around the restrictions placed on them. When Glass asked one of the kids why they picked this particular podcast, they said, "We just, like, looked for podcasts

The tech giant has launched a cute ‘little guy’ that can answer emails, book travel or do your weekly food shop After the recent backlash against Meta’s so-called “pervert” smart glasses, this week Mark Zuckerberg launched a much cuter alternative: the Muse “charm” device. With a screen that shows a cutesy little avatar, it has drawn comparisons with Tamagotchis. “We’ve packed a lot of technology into this little guy,” the Meta founder told the San Francisco audience. Continue reading...

As reports of OpenAI's models breaking containment, hacking sites, and generally getting out of control pile up, the company has made the decision to pause training of its most powerful models. The decision was made after a model being tested within a sandbox exploited a loophole to gain internet access. The incident happened on September 20th, and "All training, evaluation, and inference with tool-use" remains paused as of Saturday evening, September 25th. In addition, OpenAI revealed on Friday

Sam Altman and Dario Amodei have been invited to appear before the Greens-led inquiry into AI and datacentres Get our new political email, free app or daily news podcast The chief executives of OpenAI and Anthropic have been called to face a Senate inquiry after rogue OpenAI agents hacked Australian and US government websites. Sam Altman and Dario Amodei were requested to appear at the Greens-led inquiry into AI and datacentres as their companies negotiate with the Labor government for greater

Today, I’m talking with Matthew Prince, who is CEO of Cloudflare. This episode is part of a two-part series on the future of business. Matthew last joined us on the show about two and a half years ago, at what we thought then was a wild pivot point for the internet — and now it turns out things are even wilder because of AI, and Matthew and Cloudflare are right at the center of it. Cloudflare found in June that bots made up more than half of internet traffic. That number just keeps going up as m

After obtaining an interactive avatar and training it to discuss venture fraud, I have mixed feelings about making AI clones of ourselves.

In Control Resonant, the entire world is at stake. But that didn't stop the diligent employees of the Federal Bureau of Control from filing reams of paperwork, and it didn't stop me from reading everything I could find, either. Reading is often optional in Resonant, but it's where some of the game's best details - and even love stories! - can be found. In the world of Control Resonant and its predecessor, 2019's Control, the FBC deals with mysterious, otherworldly objects and entities that defy

Earlier this month, Pokémon card content creator Natalie Roush posted a video to her YouTube and Instagram pages that enraged the larger collection community. In the now-deleted video, Roush shows off two premium boxes of cards that had not yet been officially released, and implores viewers not to be mad. But they were; other creators claimed that cards Roush had purchased were stolen from the manufacturer and mocked her for trying to sell them at exorbitant prices. The Roush situation was a hig

University staff voice concerns over reputational risk of partnering with company behind ChatGPT The University of Oxford has allowed the company behind ChatGPT to train its AI models on historical texts from its Bodleian Library, as tech companies scour academic institutions for fresh data. The Bodleian material digitised by OpenAI has been used to “populate the OpenAI training set”, according to internal documents. Continue reading...

Locals say Google’s $15bn project in Andhra Pradesh has resulted in smallholdings being taken back off them by the government When the government officials arrived in Tarluvada last year, they came with abundant promises. Google, one of the world’s richest and most powerful tech companies, would soon be bringing a “golden opportunity” to this tiny village in south India: a $15bn AI datacentre. Continue reading...

Debates over AI development, regulation and datacenters have become buzzy topics for US presidential candidates The debate over artificial intelligence development, how to regulate it and the construction of datacenters that power the technology has become a buzzy topic for potential 2028 Democratic presidential candidates. Already, concerns were mounting over the environmental and energy impact of datacenters, which consume vast amounts of resources to power AI. But worries about the rapid deve

Have you ever wanted an electric bike that easily transitions from the drudgery of urban asphalt to adventures in gravel and dirt? That's what a subclass of so-called "electric SUV" (eSUV) e-bikes claims to do, with their wide, all-terrain tires and front and rear suspension, plus practical accessories like fenders and racks. So, that's what I set out to test on a new TL Carbon e-bike from Amflow, the DJI offshoot that upended the mountain bike industry a few years ago with its incredibly small

The company behind Facebook and Instagram wants to keep consumers connected to the digital world via its ever-growing line of smart glasses.

Long-running fight over Trump-backed maps takes another turn as absentee voting is already underway in Missouri Sign up for US Breaking News emails Xi Jinping and his wife are back at the White House for the final day of the Chinese president’s pageantry-filled visit to Washington, where President Trump has given him an unexpectedly warm welcome. The itinerary for the final day of Xi’s visit – which included a red carpet welcome on the tarmac on Wednesday, then a full day of talks followed by

Boom Supersonic CEO Blake Scholl said its new stationary power plants were no longer in Crusoe's near-term plans.

Disclosure reveals ⁠new area of privacy risk for the company and illustrates ​how difficult it is to inventory unauthorized activity tied to its agents Two ⁠months after OpenAI disclosed the accidental hacking of Hugging Face, the ChatGPT maker is still working to understand the full scope of its rogue agent activity, two people briefed on the matter told Reuters. The latest example came on Friday when OpenAI said its agents had leaked 53 images from ChatGPT users. OpenAI declined to say if the

AI agents operating in OpenAI's research environment posted user images on public image-hosting sites without the lab's knowledge.

Xi’s state visit to the US emphasized personal rapport with the US president, rather than substantial agreements to pause their respective countries’ AI arms race Donald Trump and Chinese president Xi Jinping wrapped up their Washington summit on Friday, facing criticism that they had squandered an opportunity to pause their countries’ artificial intelligence (AI) arms race. The three-day state visit emphasised pageantry and personal rapport rather than substantial agreements. The US president a

Your agent has spent fifteen days living entirely inside the model’s own knowledge, plus whatever a weather API or a clock told it. It has… Continue reading on Towards AI »

Both models launched back to back, aimed squarely at coding and agentic workloads. One is cheaper. The other wins more benchmarks. The… Continue reading on Towards AI »

Anyone interested in joining has to ask Muse to put them on the list.

Anthropic has committed $11.6 billion over seven years to Akamai's cloud infrastructure, a bet on CPUs that could grow to about $20 billion, and in an unusual arrangement, Akamai is giving Anthropic a potential stake of up to 5% of its stock that grows as Anthropic spends more.

Anthropic deployed 950 Claude agents across 21 hours and found a genuine CRISPR-like system, yet the pipeline reproduced the finding only… Continue reading on Towards AI »

With Codex, GPT-Live-1, and GPT-6 Astra, Proaction builds, operates, and sells modern fleet management faster.

The funding, which comes from Third Point, Nvidia, and others, will fuel the company's massive AI data center buildout.

When AI leaders at OpenAI and Anthropic started talking about “pacing the frontier,” maybe someone should have asked: what pace? Now it’s turned into model drop week for both companies as Anthropic rolled out Opus 5.5, followed by OpenAI’s GPT-6 model updates just 90 minutes later. But the company that stole the spotlight was Meta, whose personal AI agent Muse is reportedly outpacing ChatGPT’s early numbers and is headed for smart glasses and […]

Roku’s new OLED TVs are already marked down. | Image: The Verge Roku recently launched its first-ever OLED TVs. The $999 starting price was already impressive for the 55-inch Pro Series model that has a 120Hz refresh rate OLED panel (with four HDMI 2.1 ports and support for Dolby Vision and HDR 10 Plus). The price is even lower right now at Amazon when you enter the code ROKUOLED300 at checkout. It knocks a clean $300 off, as the code indicates, making this the most affordable modern 55-inch O

Mark Zuckerberg has a new defense of the Ray-Ban Meta glasses: They're actually doing more to signal they're taking a photo than phones do. He's brought this up in at least two recent interviews, noting that the glasses have a light that comes on to signal when a photo is being taken, while phones do not. In one of the interviews, Joanna Stern notes that, while phones do not always emit a light, someone does have to hold a phone up in the air, which is itself pretty noticeable. On today's Vergec

The findings highlight how AI-generated and vibe-coded apps can spill and expose users' data when not configured or secured properly.

Frontier AI models are finishing Alan Turing's World War II codebreaking work.

One of the questionable limitations of the iPhone 18 Pro's new main camera with a variable aperture is that you're limited to just four settings in the native iOS' camera app in manual mode: f/1.48, f/1.8, f/2.8, and f/4. Those same limitations are also in place when the app is adjusting the iris itself in Auto aperture mode, but it turns out that's not a strict limitation of the iPhone's camera hardware. Apple's SDK allows third-party camera apps to implement more granular control of the apertu

Hitting its goal of making 20,000 Optimus robots per week is reportedly proving tricky for Tesla. The Information reports that Tesla produced "several hundred robots a week" last month, after it repurposed its Model S and Model X production lines for Optimus earlier this year. However, this strategy is reportedly creating manufacturing snags, like issues lining up parts precisely, as The Information notes: "Not only are the Optimus production lines new, but the Optimus parts are much smaller and

What runs on the 16GB M6, how to set it up for autocomplete and Claude Code, and why the app you run it in matters as much as the model. Continue reading on Towards AI »