TheChek News Flow

DAX:

MDAX:

SDAX:

Nous Research, the startup behind the open-source Hermes agent, is finalizing a new round of funding led by Robot Ventures, with significant participation from USV and other prominent investors at a $1.5 billion valuation, according to three sources with knowledge of the deal. The company is raising at least $75 million, and fielded a high level of interest from investors, according to the people.

Nous Research declined to comment. USV and Robot Ventures didn’t respond to our request for comment.

The company was founded in 2023 by Jeffrey Quesnelle, Karan Malhotra, Ryan Teknium, Shivani Mitra. Before this round, it had raised a total of $70 million in funding from investors including Paradigm, Robot Ventures, North Island Ventures, OSS Capital, and Balaji Srinivasan, according to Crunchbase.

Weeks after Openclaw’s agent went viral, Nous Research released its own competitor called Hermes. OpenClaw is an agent that runs locally on a PC and can perform tasks on behalf of the user. One key difference is that Hermes shipped with built-in “skills,” such as web search, coding and image understanding. Furthermore, it was designed to automatically learn from people’s usage and build more skills without manual intervention. Additionally, the startup has released language models focused on coding and math.

Just like Openclaw, users can automate tasks with Hermes and chat with these agents or receive messages from them in apps like Telegram and Discord. These tools have become increasingly popular as they allow users to run their AI agents remotely and around the clock.

Open-source and widely adopted, Hermes has amassed a massive following on GitHub, boasting roughly 214,000 stars and nearly 40,000 forks. Developers can run Hermes on a desktop or on a virtual private server.

But Nous Research also offers a cloud-hosted version, which some users may find to be more user-friendly, avoiding any setting up on their own machines. The hosted version is available via various paid tiers ranging from $20-$200 a month.

Sources say the new funding will help to expand Hermes’ products and business model further.

Unrelenting demand for computing power has investors looking for chances to own the physical infrastructure behind AI

America’s data-center developers are ready to cash in on the AI boom.

Data-center builders and operators across the U.S. are working with bankers to sell majority equity stakes worth tens of billions of dollars in their companies this summer, according to people familiar with the efforts.

Bankers are working to sell stakes in firms including Netrality Data Centers, DataBank, Edged, EdgeCore Digital Infrastructure and others with properties located from Phoenix to Atlanta. They are pitching private-equity firms on a hot asset class benefiting from unrelenting demand for computing power.

Sales of data-center operators are on the rise as owners of these firms seek exits and investor interest in owning the physical infrastructure behind advanced artificial intelligence grows. A massive backlog of demand for server capacity pushed companies to pursue novel strategies to secure more of it, from renting chips from direct competitors to launching data centers into orbit.

Shortages of everything from electricians and plumbers to gas turbines and memory chips have driven the cost of building new data centers ever higher. Nvidia Chief Executive Jensen Huang recently estimated the cost of building a gigawatt of new computing power using his company’s architecture could soon reach $80 billion to $100 billion. Some operators have been forced to seek deeper-pocketed backers to fund their next phase of expansion, according to several people working on the deals.

Among the largest potential deals in the works is the sale of a majority stake in Dallas-based DataBank, which could reach as much as $25 billion, according to people with knowledge of the process. DataBank is owned by a consortium led by DigitalBridge that also includes Swiss Life, EDF Invest and Australian pension fund AustralianSuper.

EdgeCore Digital Infrastructure, a Denver-based data-center developer and operator of data centers for large cloud providers, asked potential buyers to submit offers last week, according to a person with direct knowledge. EdgeCore is owned by Partners Group, a Swiss private investment firm that has poured billions into the company to support expansions in Virginia, Arizona and other markets.

The potential stake-sale talks are at varying stages, and some might not result in sales, the people said. KKR’s newly formed Helix Digital Infrastructure is among those evaluating the assets, according to a person with direct knowledge, in addition to other firms.

The current push builds on a record. In 2025, data-center M&A hit about $50 billion, more than double the prior year, according to S&P Global Market Intelligence.

Whether there are enough buyers with the means to absorb so many large deals at once this summer remains an open question. Ravi Purohit, the co-head of infrastructure at Paul Weiss, said that while plenty of investors want exposure to data centers, only a handful of firms can afford multibillion-dollar deals.

“There are not that many people who can write those checks,” Purohit said.

Bankers and investors involved in data-center sales this summer say guaranteed access to electrical power is a major factor in deal valuations. Concerns around rising costs and delays might also drive some buyers to tie some of their payments to completion of milestones, they said.

Private-equity firms have increasingly added data-center operators to their portfolios.

Blackstone took a roughly 49% stake in Rowan Digital Infrastructure, a hyperscale developer backed by Quinbrook Infrastructure Partners, in April. The private-equity giant also bought data center developer QTS for around $10 billion in 2021.

Last year, a consortium led by BlackRock’s Global Infrastructure Partners and MGX agreed to buy Aligned Data Centers for about $40 billion, the largest data-center deal ever. The acquisition is expected to close in the coming weeks. The prior record was Blackstone’s roughly $16 billion acquisition of AirTrunk in 2024.

But as these megadeals scale up, developers are increasingly running into fierce local opposition, driven by residents’ anxieties over rising utility bills, noise and advanced AI services in general. In some cases, it has pushed developers to pause or walk away from projects. This all makes investing in data centers riskier than it was in the past, investors said.

“I think people are downplaying the significance of Nimbyism in this country right now,” Purohit said, adding that companies looking to sell this summer will be scrutinized for how they plan to handle it.

“The more they can demonstrate to buyers that they have a constructive relationship with these communities…that actually goes a really long way,” he said.

The social-media giant’s AI-enabled, camera-equipped smartglasses have become a privacy sore spot

Meta META -1.86%decrease; down pointing triangle is betting that AI-enabled smartglasses capable of capturing everything you see and hear are the future—and the company is doing all it can to make its fast-evolving devices catch on.

Company executives wear smartglasses at public events. Meta donated more than 100,000 pairs to blind veterans on President Trump’s birthday weekend. And it has partnered with celebrities and creators, including Kylie Jenner, for help promoting them.

Along the way, the camera-equipped, audio- and video-recording devices have become a privacy lightning rod.

Meta employees were asked to take them off at a trial where Chief Executive Mark Zuckerberg was testifying earlier this year. On social media, people have complained about being filmed without their knowledge, with accounts surfacing of users disabling the white light that signals the glasses are recording. (Meta recently addressed the problem with a software update.)

The glasses—made in partnership with eyewear giant EssilorLuxottica—are part of Meta’s grand plan to control the next phase of the internet. The company is working to bolster the revenue it earns beyond ads and build a future in which artificial-intelligence assistants help us live and work.

It envisions a future in which consumers still have smartphones, but increasingly use other devices worn on their bodies to talk to AI chatbots, send texts and take pictures or videos. Meta also is experimenting with different AI device form factors, including small devices that users could either wear as a pin, don as jewelry or put in their pockets, according to people familiar with the matter.

Meta has said glasses are likely to be the new dominant form factor for non-smartphone AI devices, in part because of what the company believes is a large addressable market. Nearly two billion people already wear prescription glasses, Meta executives frequently say.

Meta has launched different versions of the product in the past few years: the $499 Oakley sports glasses that offer AI-powered motivation and fitness tracking, $379 Meta Ray-Bans, and non-Ray-Ban glasses for $299. The less expensive glasses are meant to help Meta get its devices into the hands of as many people as possible.

At a recent event, Meta unveiled a pair of sleeker, oval-shaped glasses designed in collaboration with Jenner. The beauty brand founder took selfies with Meta executives at the New York event while wearing the shades. Fashion and tech influencers have also been hawking the glasses on TikTok and Instagram.

The company plans to release a new generation of the Ray-Ban smartglasses at its annual developer conference in September, according to people familiar with the matter. The company sold seven million pairs of the glasses in 2025.

“We’re at this moment now where everyone had flip phones and then smartphones came out, and it felt pretty clear that in five years or whatever, all of the flip phones were going to be smartphones, and that’s basically how I feel about glasses today,” Zuckerberg said in a recent interview with Complex.

Meanwhile, Meta has been testing two features for its glasses that have drawn the ire of privacy advocates. The first feature, called NameTag, would remember the faces of people that the user met and remind them who the person was the next time they ran into them.

Meta’s technology chief, Andrew Bosworth, said on a recent podcast that a user would say “remember this person” about someone they met in real life. The data would only be available to the user when they are wearing their glasses, would be encrypted locally to the device and wouldn’t be stored in a central database, he said.

More than 70 local, state and national organizations that advocate for consumer privacy and civil rights signed a letter in April, calling on Meta to halt and publicly disavow its plans to equip its glasses with the NameTag feature.

“The principle here is quite simple: Your glasses should not know my name,” Cody Venzke, a senior staff attorney for the American Civil Liberties Union, said in a news release.

Meta says it hasn’t yet shipped the feature to consumers and no final decision on whether to do so has been made.

Another feature it is exploring would allow the glasses to constantly capture audio and visuals without notifying those around the user via the white light, according to people familiar with the project. The data captured from that feature wouldn’t be stored by Meta or made available to the user directly via a library of audio and video, one of the people said.

Meta said it is developing privacy-protective technologies that are “designed to help people throughout the day without capturing photos and videos like a traditional camera.”

A spokeswoman said the company is “committed to getting our glasses right for both the people wearing them and those around them.”

Meta also recently filed a patent application for a system that would record a user throughout the day to assess their mood and create customized workout plans for them based on that data, according to a public filing.

To illustrate the concept, the patent application gives examples of what Meta might record.

“User laughs with friend at dinner at 5:15 p.m. Audio is recognized and logged by AI,” the filing says. “User sighs at 9:15 p.m. AI is listening from a smart home device and logs it.”

A Meta spokeswoman said the company files patent applications to explore inventions or ideas but filing one doesn’t necessarily mean it is actively developing the technology.

The Financial Times previously reported some details of the new Meta glasses feature and The Information earlier reported on the company’s work on a potential AI pendant.

Across the country, some locales are starting to ban smartglasses. Earlier this month, the New York state court System sent out a memo, prohibiting anyone in their courthouses from wearing the devices.

DFSX’s investors include state-backed entities and industrial funds such as a venture-capital vehicle co-founded by Alibaba’s Jack Ma

A Chinese startup has stepped out of the shadows with the release of an artificial-intelligence chip, aiming to prove that China can power its AI boom entirely on homegrown technology.

Dongfang Suanxin, or DFSX, introduced its self-designed AI chip—the DF1000—on Monday in Shanghai, alongside a product roadmap promising a second-generation chip by the end of this year and a third iteration in 2027, it said in a social-media post.

According to materials published by DFSX, it is using a fully domestic supply chain to bypass restrictions on high-end technology that have hindered Chinese firms’ ability to produce cutting-edge semiconductors.

Founded in 2024, little has been known about DFSX until recently, a decision the company said it made to “keep its head down and focus on core breakthroughs.”

According to Chinese corporate registry data, DFSX is valued at roughly $1.8 billion following its latest financing. Investors include state-backed entities and industrial funds such as a venture-capital vehicle co-founded by Alibaba’s 9988 -0.90%decrease; red down pointing triangle Jack Ma.

While major players like Nvidia chase raw power with tiny transistors and expensive, highly restricted High Bandwidth Memory, DFSX uses a physical workaround. By stacking custom memory on top of the computing layer like Lego bricks, it eliminates data traffic jams—delivering memory bandwidth without foreign-made HBM.

Designed to handle heavy AI training and inference workloads, DFSX says its products overcome three bottlenecks: It doesn’t rely on cutting-edge manufacturing, breaks through memory bandwidth walls and features a full-stack software ecosystem.

The startup uses a legacy 14-nanometer manufacturing, but claims the DF1000 can match some mainstream Western chips built on far more advanced 4-nanometer processes in certain inference workloads. Though it still trails in training, DFSX says its next-generation DF2000 will close that gap.

Led by Wei Shaojun, a senior central government adviser and professor at Tsinghua University in Beijing, the company’s design marks a departure from the scale-out strategies of chip giants like Huawei Technologies, which compensate for outdated manufacturing by clustering tens of thousands of chips into supernodes.

DFSX is instead commercializing 20 years of Tsinghua research by placing data storage right next to the processor’s brain to eliminate delays, allowing the chip to reshape its internal wiring to match the software it is running.

Tech policy experts say success would prove that Chinese companies can build competitive AI despite restrictions, disrupting a Western-led market.

DFSX’s software-defined architecture aligns with a broader shift from raw horsepower to specialized efficiency. David Bennett, chief executive of Japanese sovereign AI firm ai&, says the transition to application-specific hardware is the industry’s defining trend.

“We can see the macro trend playing out,” he said.