Category: 3. Business

Meta CEO Mark Zuckerberg is shaking up the artificial intelligence domain with a massive, unprecedented recruitment drive that is not only targeting OpenAI’s top talent but also making it even harder for the US government to build its own tech bench.

Zuckerberg is offering mind-boggling compensation packages, sometimes exceeding $100 million in the first year alone to lure leading AI researchers from OpenAI and other companies. Over four years, total payouts could soar to $300 million, as reported by WIRED. These are not just high salaries, they rival the kind of money usually reserved for star athletes or major start-up valuations.

The campaign culminated this week with Zuckerberg unveiling Meta Superintelligence Labs (MSL), his new elite AI division. The Meta founder has personally courted potential hires at his residences in Palo Alto and Lake Tahoe. His most high-profile recruit so far is Alex Wang, co-founder of Scale AI, who will serve as Meta’s Chief AI Officer. Former GitHub CEO Nat Friedman will lead product and applied AI development. Eleven other top-tier hires were listed in an internal memo.

Zuckerberg’s all-out raid is dramatically inflating AI compensation and intensifying a talent war already underway in Silicon Valley. The ripple effect is particularly damaging for the U.S. government, which was already struggling to compete for AI expertise. With tens of millions now easily attainable in private industry, public service becomes a much harder sell.

Meanwhile, Chinese tech firms are quickly gaining ground, supported by their government’s ability to direct top talent into state projects. A recent Wall Street Journal report warned that China’s AI models from companies like DeepSeek and Alibaba—are rapidly gaining traction across Asia, Europe, and Africa.

Backdrop and implications

Zuckerberg’s hiring spree is part of a larger strategic pivot, reminiscent of his earlier move to shift Facebook’s focus to mobile. Then, he bought Instagram and WhatsApp to catch up. Now, instead of acquiring companies, he’s betting on individuals.

It’s a bold move—but not without risk. Meta has spent heavily developing its large language model, Llama, to catch up with ChatGPT, Claude, and Gemini. But The Wall Street Journal notes that Meta’s track record in generative AI has made some recruits hesitant.

Still, Zuckerberg sees the opportunity clearly. OpenAI has projected massive growth—$10 billion in annual revenue already, with targets of $125 billion by 2029 and $174 billion by 2030. Anthropic, another OpenAI spinoff, is on a $4 billion annual revenue pace. For Meta, the payoff of dominating this sector could be trillions in long-term gains.

Altman’s response

OpenAI CEO Sam Altman acknowledged the aggressive poaching attempt, telling employees that Meta did manage to hire “a few great people” but largely missed out on OpenAI’s top talent. In a Slack message, he commented, “Missionaries will beat mercenaries,” stressing that OpenAI’s strength lies in its mission-driven culture.

He also pointed out on a recent podcast that OpenAI’s financial model rewards success with strong long-term incentives, aligning innovation with economic gain.

The broader concern

This highly public bidding war reflects an underlying AI arms race that’s now impacting national interests. For government agencies, the challenge is existential. They’re increasingly priced out of a market where the world’s biggest corporations treat top researchers like venture-backed unicorns. And without major reforms or incentives, Uncle Sam may be left watching from the side-lines.

LIVINGSTON, N.J., July 3, 2025 /PRNewswire/ — CoreWeave (Nasdaq: CRWV), the AI Hyperscaler™, today announced it is the first AI cloud provider to deploy the latest NVIDIA GB300 NVL72 systems for customers, with plans to significantly scale deployments worldwide.

The NVIDIA GB300 NVL72 represents a major leap in performance for AI reasoning and agentic workloads, delivering up to a 10x boost in user responsiveness, a 5x improvement in throughput per watt compared to the previous generation NVIDIA Hopper architecture, and a 50x increase in output for reasoning model inference.

“CoreWeave is constantly working to push the boundaries of AI development further, deploying the bleeding-edge cloud capabilities required to train the next generation of AI models,” said Peter Salanki, Co-Founder and Chief Technology Officer at CoreWeave. “We’re proud to be the first to stand up this transformative platform and help innovators prepare for the next exciting wave of AI.”

CoreWeave collaborated with Dell, Switch, and Vertiv to build the initial deployment of NVIDIA GB300 NVL72 systems, enabling greater speed and efficiency to bring the latest accelerated computing offerings from NVIDIA to CoreWeave’s AI cloud platform.

The deployment of GB300 NVL72 is tightly integrated with CoreWeave’s cloud-native software stack, including its CoreWeave Kubernetes Service (CKS) and Slurm on Kubernetes (SUNK) to its deep observability and custom-designed Rack LifeCycle Controller (RLCC). CoreWeave recently announced that hardware-level data and cluster health events are now integrated directly through Weights & Biases’ developer platform, which CoreWeave acquired earlier this year.

This achievement continues CoreWeave’s legacy of delivering first-to-market access to the world’s most advanced AI infrastructure demanded by the world’s leading AI labs and enterprises. This initial deployment of NVIDIA GB300 NVL72 rack-scale systems expands on CoreWeave’s existing Blackwell fleet, which also includes the NVIDIA HGX B200 and the NVIDIA GB200 NVL72 systems. Last year, CoreWeave was among the first to offer NVIDIA H200 GPUs and was the first AI cloud provider to make NVIDIA GB200 NVL72 systems generally available.

In June 2025, CoreWeave, in collaboration with NVIDIA and IBM, submitted the largest-ever MLPerf® Training v5.0 benchmark using nearly 2,500 NVIDIA GB200 Grace Blackwell Superchips, achieving a breakthrough result on the most complex model, Llama 3.1 405B, in just 27.3 minutes. CoreWeave is the only hyperscaler to achieve the highest Platinum rating by SemiAnalysis’s GPU Cloud ClusterMAX™ Rating System, an independent AI cloud industry benchmark.

About CoreWeave
CoreWeave (Nasdaq: CRWV), the AI Hyperscaler™, delivers a cloud platform of cutting-edge software powering the next wave of AI. The company’s technology provides enterprises and leading AI labs with cloud solutions for accelerated computing. Since 2017, CoreWeave has operated a growing footprint of data centers across the US and Europe. CoreWeave was ranked as one of the TIME100 most influential companies and featured on Forbes Cloud 100 ranking in 2024. Learn more at www.coreweave.com.

Forward-Looking Statements
This press release contains forward-looking statements within the meaning of the Private Securities Litigation Reform Act of 1995 that involve risks and uncertainties, which statements are based on current expectations, forecasts, and assumptions and involve risks and uncertainties that could cause actual results to differ materially from expectations discussed in such statements. These forward-looking statements are only predictions and may differ materially from actual results due to a variety of factors including, but not limited to, general market, political, economic and business conditions. These factors, as well as others, are discussed in CoreWeave’s filings with the Securities and Exchange Commission, including the sections titled “Special Note Regarding Forward-Looking Statements” and “Risk Factors” in CoreWeave’s Quarterly Report on Form 10-Q for the quarterly period ended March 31, 2025. All forward-looking statements contained herein are based on information available as of the date hereof and CoreWeave does not assume any obligation to update these statements as a result of new information or future events.

SOURCE CoreWeave

By Marcela Ayres and Bernardo Caram

BRASILIA (Reuters) -The BRICS group of developing nations is set to announce a new guarantee fund backed by the New Development Bank (NDB) to lower financing costs and boost investment, two people familiar with the matter told Reuters.

The initiative, modeled on the World Bank’s Multilateral Investment Guarantee Agency (MIGA), aims to address global investment shifts amid uncertainty surrounding U.S. economic policy, the sources said on condition of anonymity.

Brazilian officials view the fund as the centerpiece of the BRICS financial agenda during the country’s rotating presidency. The fund is expected to be mentioned in the joint statement at the BRICS summit in Rio de Janeiro next week, said the sources.

Originally formed by Brazil, Russia, India and China, the BRICS group later added South Africa and recently expanded to include other developing nations to increase its influence in global governance.

The proposed BRICS Multilateral Guarantee (BMG) mechanism, incubated within the NDB, has received technical approval from member states and awaits final signoff from BRICS finance ministers, considered a formality, one of the sources said.

Brazil’s Finance Ministry declined to comment on the matter.

The initiative will not require additional capital from member countries at this stage. Instead, it aims to channel existing NDB resources to projects in developing nations.

No initial funding value has been disclosed, but officials involved in the talks expect each dollar in guarantees provided by the NDB to mobilize between five and ten dollars in private capital for pre-approved projects.

“This is a politically significant guarantee instrument. It sends a message that BRICS is alive, working on solutions, strengthening the NDB and responding to today’s global needs,” one source said.

Technical preparations setting up the fund are expected to conclude by the end of this year, paving the way for pilot projects to receive guarantees in 2026.

BRICS countries face challenges common to developing nations in attracting large-scale private investment in infrastructure, climate adaptation and sustainable development.

Officials argue that guarantees issued by the NDB, whose credit rating is higher than that of most member countries, could help mitigate perceived risks for institutional investors and commercial banks.

(Reporting by Marcela Ayres and Bernardo CaramEditing by Manuela Andreoni, Brad Haynes and Louise Heavens)

WASHINGTON — The European Space Agency will soon select the finalists for a competition intended to support the development of new launch vehicles by European companies.

During a panel discussion at the Paris Air Show June 17, ESA Director General Josef Aschbacher said that the agency received 12 proposals for the European Launcher Challenge, a program to award launch contracts to new vehicles as well as fund demonstrations of upgraded vehicles. Companies are eligible for up to 169 million euros ($199 million) each.

Those 12 proposals are currently going through technical reviews by ESA, which will select a group of them for funding consideration at the ESA ministerial conference in late November.

“It will be that not all 12 of these proposals will go the ministerial,” he said. “I cannot predict how many will be left after this evaluation period and therefore how many will go to the ministerial.”

ESA is using an alternative approach to funding the European Launcher Challenge than its traditional georeturn approach, where member states subscribe to programs and are guaranteed contracts in amounts proportional to the funding they provide. Instead, ESA plans to select a group of companies, after which member states at the ministerial will determine which ones they want to fund.

“Quite a few member states are preparing their decisions in case their candidates are selected as being brought forward for funding at the ministerial conference,” he said. “If you ask me the number that will go to the ministerial, it’s too early to say, but we should know in a couple of weeks.”

Toni Tolker-Nielsen, ESA’s director of space transportation, said after the panel that he expected the downselected companies to be announced as soon as July 7.

ESA has not disclosed the companies that did submit European Launcher Challenge proposals, although some companies widely believed to be participating have been announcing milestones in recent weeks to emphasize the progress they are making.

MaiaSpace, a French company developing a launch vehicle with a reusable first stage, hosted several French government ministers at its facilities outside Paris June 13. There, the company announced its intent to build a 10,000-square-meter factory there for producing the vehicle.

Yohann Leroy, chief executive of MaiaSpace, said at the Paris Air Forum earlier the same day that there was demand for small launch vehicles despite competition from rideshare services that have stymied other vehicle developers. Dedicated small launchers are often compared to taxis while rideshares are linked to buses.

“The market for microlaunchers exist. There are a lot of people who are interested in a taxi, provided the taxi can be the price of the bus,” he said. “If you want to succeed in launching a taxi, you have to make the price close to the price of the bus.”

Some companies have hinted they are interested in using the European Launcher Challenge to support work on larger vehicles. “Microlaunchers can never compete in price per kilo” against larger vehicles, said Miguel Bello Mora, chairman of the board of Orbex, said at the same Paris Air Forum panel.

That company is working on its Prime small launcher but has announced plans for a larger vehicle, Proxima, even before the first Prime launch. “We believe there is a gap and there is room for several players,” he said. “Medium size is where we target.”

Not everyone in the European small launch vehicle industry is satisfied with the competition. “It’s a pretty weird program,” said Stanislas Maximin, executive chairman of Latitude, in an interview. He said he felt the competition was either a way to help companies already far advanced in technology and fundraising or those who have struggled to raise private financing.

Nonetheless, he said Latitude submitted a proposal for the competition. “It will help us improve the launch system,” he said, such as increasing the payload capacity of its Zephyr rocket from 200 to 300 kilograms. “It allows us to go faster.”

Latitude, based in France, is also using the competition to expand its presence in Europe. “What it helps us to do is be more European. We’re building relationships with European partners,” he said.

While he said he was concerned the program will be used by some countries to support companies that don’t necessarily need such assistance, “we feel like we are in a good position to win it.”

Related

Changes to Companies House rules that would have axed an exemption allowing smaller businesses to file abbreviated accounts have been delayed amid concerns they could burden them with more red tape.

Legislation brought in by the previous Conservative administration is due to compel companies with a turnover under £10.2m, balance sheets under £5.1m and fewer than 50 employees to disclose more detail in their annual accounts from April 2027.

However, the business secretary, Jonathan Reynolds, is understood to be have reversed the move to reduce regulatory hurdles for small firms. A government source said: “We have paused them, Jonny is worried it’s too burdensome”.

Last month the government unveiled its industrial strategy in which it said it aims to “reduce regulatory burdens and speed innovation [by] cutting the administrative costs of regulation for business by 25%”.

The proposals to make companies publish more comprehensive financial information were designed to remove longstanding exemptions that allowed small firms to file “abridged accounts” and counter criticisms that the more relaxed rules for qualifying businesses had enabled fraud.

The intention had been that small companies would be required to use a set format to electronically file a profit and loss statement, which would be able to be viewed by the public and would include information such as turnover and costs.

Business groups had criticised the move saying that small companies would be forced to pay to use expensive computer software to meet the new regulations. However, Companies House had said the measures, which were set out in the Economic Crime and Corporate Transparency Act, would “be a critical step towards improving the quality of the data on the register”.

Companies House has long faced criticism of how it polices the information that it holds. In April it emerged that the agency had collected just £1,250 in fines after being given new powers to crack down on corruption.

A spokesperson for the Department for Business and Trade said: “This government is committed to avoiding undue burdens on businesses as part of our plan for change.”

Qantas has said it will beef up its security and threat detection in the wake of a cyber-attack affecting up to 6 million customers, as Australia’s privacy watchdog has warned attacks using social engineering to gain access to data are on the rise.

In an update to customers on Thursday, the airline said more security measures would be put in place after cybercriminals were able to gain access to a third-party system used by a Qantas airline contact centre to steal customers’ personal information.

“We’re … putting additional security measures in place to further restrict access and strengthen system monitoring and detection,” the company said.

Qantas began emailing affected customers on Wednesday evening, but had not indicated as of Thursday afternoon whether any compensation would be provided to those who had their personal information compromised.

Cybersecurity analysts indicated to Guardian Australia that, as of Thursday afternoon, the data had not yet been posted on forums or dark web locations that attackers commonly frequent.

The alleged culprit of the attack has yet to be identified but has similarities to a ransomware group known as Scattered Spider. The group has targeted airlines in the US in recent weeks by engaging in what are called social engineering attacks, or “vishing”. They involve calling the IT support for large companies, often impersonating employees or contractors to deceive IT help desks into granting access and bypassing multi-factor authentication.

An Office of the Australian Information Commissioner (OAIC) report on data breaches, released in May and covering the second half of last year, noted a rise in the number of social engineering attacks resulting in data breaches in Australia. The attacks made up 28% of all reported breaches resulting from malicious or criminal attacks.

The OAIC noted at the time that the “significant increase” was particularly significant within Australian government agencies, which reported 60 out of the 115 breaches of that kind – a 46% increase on the previous six months.

Google’s threat intelligence report in recent months has also warned of multiple threat actors using these methods to get into companies’ systems.

In a June update, Nick Guttilla, from Google’s Mandiant threat intelligence, said threat actors first build up intelligence on their target, reviewing employee positions and titles, information about their networks, cloud and email providers, and searching for publicly exposed documentation.

Some of this information can be found on company websites, as well as social media like LinkedIn.

From there, threat actors may test the IT service desk, which would routinely deal with a high volume of calls from staff needing help on password resets. According to Guttilla, attackers will see how far they can get before a staff member requests ID verification, feigning ignorance of the process to see if the staff member will relent and forgo normal procedure.

skip past newsletter promotion

after newsletter promotion

An attacker may also pretend their phone is unavailable and that they need urgent account access.

In some attacks, they persuade an employee to install an application that helps exfiltrate the data from a system quickly. It is unknown at this stage if this is what happened in the Qantas breach.

Guttilla said training staff to rigorously perform ID checks on all calls, particularly for privileged accounts with more systems access, was critical.

The minister responsible for cybersecurity, Tony Burke, did not confirm whether the Qantas attack was associated with the Scattered Spider group, but said he had been briefed and would allow the cybersecurity agencies to make announcements on any alleged culprits.

“The reality is with these networks, they’ll go where they can find vulnerability,” he said.

Burke said when companies relied on third parties for their systems, it made their cybersecurity obligations “more complex”.

The Australian Signals Directorate was approached for comment.