Anthropic is launching Claude Tag in research preview, an “always-on Claude” that lives inside Slack and acts as a persistent AI teammate. The feature lets users tag @Claude to get insights in conversations and assign tasks. It is available to Claude Enterprise and Claude Team customers starting today.

Claude Tag is an evolution of Anthropic’s existing Slack integrations, which already let users direct-message Claude or tag it in channels for on-demand help. Claude Code in Slack can also route coding tasks from channel mentions to full coding sessions on the web, posting updates back into the thread. But Claude Tag adds a layer of persistent context and memory that the previous tools could not maintain.

“As Claude follows along with its channel, it learns ever more about the work,” Anthropic said in a statement. Claude can also automatically gather facts from elsewhere in the organisation if granted permission to read other channels. The result is an AI that accumulates institutional knowledge over time rather than starting from scratch with every interaction.

The 💜 of EU tech

The latest rumblings from the EU tech scene, a story from our wise ol’ founder Boris, and some questionable AI art. It’s free, every week, in your inbox. Sign up now!

Everyone in a given Slack channel can access a single Claude identity, meaning anyone can see what Claude has been working on and pick up the conversation from where the last person left off. System administrators specify which tools, information, and channels each Claude identity can access, and each identity stays scoped to whichever channels the admins define. A Claude set up for legal work cannot seed memories into the engineering channel, for example.

When assigned a task, Claude Tag breaks it into stages and works through them using whichever tools it has access to, responding in a Slack thread with what it has created. But the more notable feature is an ambient mode that proactively jumps into conversations to keep teams updated, flag relevant information from across the organisation, and follow up on threads or tasks that have been forgotten.

The ambient mode is the feature that separates Claude Tag from a conventional chatbot. Rather than waiting to be asked, Claude monitors the channels it has been assigned to and intervenes when it judges that a team would benefit from a reminder, a summary, or a piece of context pulled from another part of the company. Anthropic says this makes it feel like “working with a real colleague, one that can produce work in public view, with far greater context and understanding than before”

Organisational context is increasingly the battleground for enterprise AI. Microsoft has been building Work IQ, an intelligence layer expressed through Copilot that draws on Microsoft Graph to understand roles, collaboration patterns, and organisational structure. Startups like Viktor have raised tens of millions to put AI coworkers directly inside Slack and Teams.

Glean, which recently surpassed $300 million in annual recurring revenue at a $7 billion valuation, is building a permissions-aware knowledge graph that sits between the model and enterprise data.

Claude Tag is Anthropic’s answer to the same problem, but its approach is narrower and arguably more disciplined. Rather than building a horizontal intelligence layer across every enterprise application, it plants the AI inside the one surface where most knowledge work already happens: the team chat. The bet is that persistent presence in Slack, combined with cross-channel memory and admin-controlled scoping, is enough to accumulate the institutional context that makes an AI agent useful.

The privacy implications are substantial. An always-on AI that follows along with workplace conversations and autonomously decides when to intervene will face scrutiny from both employees and compliance teams. Anthropic’s admin-scoping controls are the structural answer to that concern, but the real test will come when enterprise customers deploy it at scale and discover how workers respond to an AI that is always listening.

Anthropic says it is working to bring Claude Tag to other platforms in the coming weeks. For now, Slack is the only surface, which limits the feature’s reach but also constrains its complexity, a deliberate trade-off for a research preview.

HPC

Use of Arm cores and Linux mean Beijing hasn’t broken away from the world

The TOP500 list of Earth’s mightiest supercomputers has a new leader: the 2.198 Exaflop/s LineShine machine housed at the National Supercomputer Center (NSC) in Shenzhen, which took the top spot without using any kit from Nvidia, Intel, or AMD.

Which is not to say that LineShine is an entirely Chinese creation. As explained in a pre-press paper, the machine’s LX2 processors are a local effort but use Armv9 designs – so chalk up a win for Blighty, the home of Arm. The machine also runs KylinOS – a Linux distribution that features contributions from around the world.

The paper reveals that LineShine comprises 20,480 computing nodes, and that each LX2 processor “integrates two compute dies (304 cores total) and eight on-package HBM stacks (32 GB, 4 TB/s aggregate bandwidth).”

“Each compute die contains 152 cores and 128 GB of off-package DDR memory organized into four NUMA domains,” the paper adds. “A dedicated SDMA engine handles data movement between DDR and HBM. The LX2 supports FP64/FP32/FP16/INT8 via SME and SVE units, delivering up to 60.3/120.6 TFLOPS in FP64/FP32. Nodes are interconnected via the LingQi high-speed network with a dual-plane multi-rail fat-tree topology, offering 1.6 Tb/s bandwidth per node.”

That network is also a Chinese creation, from the minds at Hangzhou LingQi Technology Co.

LineShine became the first system on the TOP500 to exceed two exaflops of sustained double-precision performance using CPUs only and the curators of the list think it could do better in future tests, because this time around it reached about 80 percent of its 2.736 Exaflop/s theoretical peak in tests conducted in preparation for this iteration of the TOP500 list.

News that LineShine topped the supercomputing charts comes as China’s government increasingly steers local organizations towards buying made-in-China tech. Beijing wants to decrease dependency on foreign products, because China has gone all-in on AI and other technologies to boost economic growth and enhance the capabilities of its military.

China’s Communist Party understands that reliance on imports can stymie those ambitions, with the USA’s ban on GPU sales to the Middle Kingdom offering ample evidence of the need to control tech supply chains. And now Beijing can point to its policies producing the most powerful single computer on the planet.

It’s conceivable that China could do even better in the future, as its GPU industry is nascent and currently producing products whose performance trails Nvidia and AMD by four or five years.

Those two paragons of US computing power, along with Intel, dominate this version of the TOP500 list – as has been the case for years. China is therefore on the march, but is a long way from global dominance.

Our sibling site The Next Platform has extensive analysis of the TOP500 list here. ®

The growing use of AI contributed to Oracle laying off 21,000 workers in a year, according to a Securities and Exchange Commission filing on Monday.

In its annual regulatory filing for the fiscal year ending May 31, Oracle said it has 141,000 full-time employees. In its 2025 filing, Oracle said it had 162,000 employees. The reported 12.9 percent reduction followed March reports of mass layoffs at the database management software company.

“[T]he adoption and deployment of AI technologies across our operations have resulted, and may continue to result, in reductions to our workforce,” the filing reads.

However, the job cuts are also tied to large capital expenditure to build Oracle’s data center infrastructure to support AI workloads.

“The majority of the initiatives undertaken by the 2026 Restructuring Plan were effected to implement our continued emphasis in developing, marketing, selling, and delivering our cloud-based offerings,” this week’s filing reads.

Oracle plans to raise $45 billion to $50 billion in 2026 to expand its Oracle Cloud Infrastructure for customers like OpenAI, xAI, AMD, Nvidia, and Meta, it said in February. About half of that funding will come through debt, with the remainder coming from equity. When Oracle announced this, investors had already been concerned about Oracle’s growing debt to fuel its AI efforts. Overall, Oracle has over $120 billion in debt, per its fiscal year 2026 earnings report.

In February, bondholders sued Oracle, claiming that they lost money because Oracle hid the need to raise its debt to build its AI infrastructure, Reuters reported.

from the dogma-isn’t-journalism dept

I recognize that when we talk about AI generally, and specifically AI in the gaming industry, there are some people out there who will simply dogmatically insist that this technology doesn’t have a place in the industry and never will. This typically comes along with two chief concerns: concerns about artistic expression if AI is used in the creation process and industry jobs being lost as a result of its use. As I’ve said in previous posts, people aren’t merely allowed these opinions, but we should be glad they’re there. If we’re going to move into the next phase of AI use in gaming, we need people to challenge its use, point out its potential negative outcomes, an spur the discussion about what it means for the art that is these games.

But journalists, including opinion journalists, are supposed to have a higher standard for topics like this. And it’s with that in mind that I say the following: Kotaku’s writing on AI in gaming is dogmatic and annoying in that it pre-judges all uses of AI in gaming in the same way I described above.

Now, I generally like Kotaku and have been a reader of the site for many years. But the pre-judgment in these articles is getting cringey. Take this recent post by Zack Zwiezen about how depressing Steam Next Fest was this year because gaming companies are using AI as a tool in their games. Zack uses a tool that informs him if a game he’s looking at comes with an AI disclosure. And this brought him great displeasure.

When I opened the main hub page for Steam Next Fest earlier today, right after my email inbox was flooded by PR messages reminding me that the event had started, I was excited to go exploring. But as I did, I started running into numerous games that feature disclaimers from the devs confirming the use of generative AI tools for various parts of the game or its marketing.

I clicked on 16 different games featured on the Steam Next Fest main hub page, and 10 of them triggered my extension and warned me of generative AI disclaimers. Of course, this will vary from user to user, as many of the hub’s slots are algorithmically driven. Still, my reaction is: Yikes. And scanning the most popular demos, I continued to run into AI warnings, so it’s not just my algo serving up slop.

Advertisement

Note that Zack knows almost nothing more about these games prior to his negative reaction about it all. He does note that some of the disclosures either come from small studios that explain that they don’t have the resources otherwise to produce these games without AI, which is the exact point I’ve been making about lowering the barrier of entry into the industry, or that the use of AI was very small and reviewed and edited by humans before anything went in the game. But seriously… that’s it. And the use of AI’s mere existence has resulted in Steam Next Fest being “depressing.”

That’s not how it’s supposed to work. Unless you believe that every use of AI in gaming is bad, that is. And if that’s your stance, you’re going to spend a lot of time being upset, because this isn’t going to stop. These are tools and tools will be used in the industry.

This isn’t a one-writer problem at Kotaku, either. Rebekah Valentine recently wrote a piece about how Epic produced a video showing how some of the art in Fortnite is made. And, yes, there is some use of generative AI tools involved. From there, Valentine discusses whether editing out undesirable contributions from AI, or mistakes, was actually saving time in the creation process. Then she notes that she’s not an artist and would believe artists if they said it did in fact save time. Then she says maybe that doesn’t matter at all because AI isn’t perfect and humans don’t always catch mistakes before something makes it into a game.

But then she questions whether any of the above matters because gaming companies have laid people off.

Or maybe none of that matters. Maybe we’re content with touches of generative AI here and there if it “helps” developers. We shouldn’t refuse the conveniences of new technology, right? Maybe this is good for the artists. Sure. But this is coming just three months after Epic games laid off 1,000 people because its leadership couldn’t keep an eye on their own balance sheet. What that tells me, in aggregate, is that Epic Games needs fewer artists to make new content for the game faster, which inevitably is going to mean tighter deadlines and more mistakes. I am not naive enough to think that Epic Games, having laid off this many folks, is going to use generative AI to give the artists that remain more generous deadlines that will allow them to reach their creative potential or fully realize their original ideas or whatever. Epic Games is concerned, first and foremost, with making money. If its leadership thinks they have a tool in hand that will allow fewer people to spit out art that’s good enough faster, they will use it.

This tendency to couple layoffs in gaming companies with the use of AI is lazy. I have no doubt that some layoffs, or slowing hiring trends, have been a result of the use of AI in some companies, some of the time. And I also understand that the gaming industry has suffered from some serious layoff numbers in recent years.

But the macro story is actually that layoffs have slowed in the industry, not sped up. They appear to have peaked in 2024, with a downward trend since then. We’ll have to see how 2026 ends up looking, but I very much believe that these claims are being made due to some very high profile layoffs at major gaming companies this year. The story of Epic’s layoffs is far more complicated than them deciding to use AI. The reasons behind Microsoft’s steady drip of layoffs and studio closures over the past few years is all the more so. And, industry wide, it needs to be remembered that there was a furious spate of acquisitions and consolidation in the market during the pandemic era of 2020 and 2021. That itself came with the dual problem of gaming companies overextending and lying about future plans when the regulators came calling about some of these acquisitions.

I’m not asking anyone to like bad uses of AI in gaming. I’m not even asking anyone to like AI in gaming generally. But dogma has no place in journalism, nor the market. Bad use of AI will produce bad games and the market will respond. Judging any and all uses as bad before you know anything else about a game is not the job of a journalist. Blaming every layoff in the industry on AI is no different than blaming piracy for every lost sale.

In short, writing about how game makers are using a tool and just shouting “It’s bad!” at it is not journalism. It’s not even really opinion journalism. It’s just lazy.

Filed Under: ai, ai in video games, journalism, video games

Companies: kotaku

B&H’s flash Deal Zone drops a blowout 15-inch MacBook Air configuration to a record-low $949, with free 2-day shipping.

For 24 hours only, pick up Apple’s last-gen M4 MacBook Air 15-inch for $949 at Apple Authorized Reseller B&H Photo. Available in your choice of Midnight or Sky Blue, this laptop originally retailed for $1,199.

Buy 15″ MacBook Air M4 for $949

The 15-inch Air comes with Apple’s M4 10-core chip, along with 16GB of unified memory, and 256GB of storage. If you want to extend your storage, we covered network attached storage (NAS) deals that deliver savings of up to $215 off.

B&H is throwing in free 2-day shipping on the thin-and-light laptops when shipped within the contiguous U.S., ensuring your order will arrive quickly so you can enjoy it right away.

This Flash Deal Zone is valid now through 8:59 p.m. Pacific Time, or while supplies last.

If you need additional ports and storage, B&H is also running a coupon special on the M5 14-inch MacBook with a 1TB SSD. The Space Black model is marked down to $1,529, which is $170 off retail.

Amazon Prime Day is here, and it’s bringing some jaw-dropping deals on TVs. Whether you want a flagship OLED TV that delivers perfect blacks or a budget-friendly Mini-LED TV that punches above its weight, there is a deal for everyone this year. I have rounded up the five best TV deals so you don’t have to dig through endless listings. Let’s get into it.

LG C5 OLED Evo (65-inch)

If you want the best of the best, the LG C5 OLED Evo is the one to beat. This 65-inch flagship OLED has dropped to around $1199, down from roughly $1399. 

You get over 8.3 million self-lit pixels that deliver perfect blacks and stunning color, even in bright rooms. It’s powered by the Alpha 9 AI Processor Gen8 chipset, which provides all the performance needed to stream content smoothly and handle anything you throw at it.

The TV comes verified for glare-free viewing, so it performs great no matter the lighting in your room. Gamers get a 0.1ms response time, up to 144Hz refresh rate, and four HDMI 2.1 ports, plus NVIDIA G-Sync and AMD FreeSync. Toss in Dolby Vision and Dolby Atmos, and you have a TV that handles everything. 

Samsung S90F OLED (55-inch)

If you want an OLED TV at a friendlier price, the 55-inch Samsung S90F will be right up your alley. It has dropped to around $997.99, down from roughly $1,397.99. That is up to 28% off and a great entry point into premium OLED territory.

This TV runs on Samsung’s NQ4 AI Gen3 processor, which uses 128 neural networks to upscale everything you watch to crisp 4K quality. You get powerful brightness, deep contrast, and smooth motion for tear-free gaming at up to 4K 144Hz. 

It can also transform SDR content to HDR-like quality with brighter highlights and more vibrant colors. If you want vibrant colors and inky blacks for mixed use and gaming, this one is a steal.

Samsung The Frame LS03F (55-inch)

The Samsung Frame is for those who want their TV to double as a piece of art. This 65-inch model has dropped to around $697.99, down from roughly $1,097.99, which is about 36% off.

When you are not watching, the Frame transforms into art with its matte, glare-free screen that makes digital paintings look like real prints. You can upload your own photos or pick from a curated collection in the Art Store, and customizable bezels let you match it to your decor. 

The slim design mounts flush to the wall, and an external hub connects the TV to power and your devices with a single wire, so you don’t have to deal with messy cables. I have a Samsung Frame TV in my bedroom (a different model), and let me tell you, it looks far better than your regular TVs. 

TCL QM7K Mini-LED QLED (55-inch)

If you are shopping on a budget, the TCL QM7K is the deal for you. This 65-inch Mini-LED QLED has dropped to $498.99, down from $649.99, which is about 23% off and lands it under $500, which makes it a bargain in TV territory. 

TCL’s QD-Mini LED combines the best of QLED and OLED tech, with up to 2500 precise local dimming zones for dark black levels. The high HDR brightness gives you a great picture in any room, and the CrystGlow HVA panel blocks reflections so your image stays crisp and visible. 

Then there’s TCL’s Halo Control System working behind the scenes to deliver clean, halo-free images without that distracting glow around bright objects. With Google TV baked in, this is incredible value for the money.

Sony Bravia XR8B (65-inch)

No best TV list can be complete without a Sony Bravia, and the one we are featuring on this list is the Sony X90L BRAVIA XR. The TV has an MSRP of $1398. But this Prime Day, the TV is getting a massive price cut, going down to $1198, giving you a savings of $200. 

As for features, it hits all the right notes. It packs a 4K OLED panel with over 8 million self-lit pixels that are precisely controlled to deliver pure blacks with high brightness, while the XR Processor enhances every scene in real time, boosting color, contrast, and clarity on the fly. 

One of my favorite features of this TV is the Studio Calibrated Picture mode, which ensures you watch your favorite movies just the way the creator intended. 

So, which TV should you buy?

It really comes down to your budget and what you want from your TV. If money is no object, the LG C5 OLED and the Sony Bravia are the clear winners. If you want OLED quality at affordable pricing, the Samsung S90F is a brilliant pick. The Samsung Frame is for the design lovers, while the TCL QM7K is the budget champion that doesn’t skimp on picture quality.