Tech
Satya Nadella has issued a shocking warning to companies using AI
Of all the debates raging about the potential downsides of AI, there is one worry causing the most hand-wringing among AI enthusiasts in Silicon Valley. Their fear is that the giant AI labs that sell proprietary models are somehow acting like Trojan horses.
The concern is that, as startups and enterprises use AI models from labs like OpenAI and Anthropic, the labs gain ever-increasing access to those companies’ most sensitive business information. The model makers can then use that knowledge for themselves, potentially becoming competitors to their own customers. Those issuing such warnings range from VCs like Jason Calacanis to Palantir CEO Alex Karp.
Now, in a surprising blog post published on Sunday, Microsoft CEO Satya Nadella has joined this crowd. Nadella warns that AI users (the “buyers” as he calls them) are paying twice. They knowingly spend for AI token usage but they also, obliviously, hand over valuable data in the process.
“You essentially pay for intelligence twice, once with money, and again with something even more valuable: the proprietary knowledge you must reveal to make that intelligence useful. The better you want the model to perform, the more of that knowledge you have to feed it!” he writes.
Most dangerously, enterprises are literally teaching the models about the nuances of their businesses, he argues.
“Models learn from ‘exhaust,’ the prompts people write, the tools agents use, and especially the corrections people make when the model is wrong. Every correction is distilled into institutional know-how,” he writes.
This is “the kind of knowledge a competitor could never buy,” and yet enterprises are handing it over.
Nadella argues that if AI companies get to freely scrape the internet to train their models, it’s only fair that enterprises get to study — or “distill” — those models in return. “Distillation” is the practice of using a model’s own outputs to learn how it works and to train a new, often cheaper, model based on those insights. In February, Anthropic accused Chinese open source models of sending millions of prompts to Claude as a way to improve their own models, and urged the U.S. government crack down on export controls.
Nadella’s point is that model makers can’t have it both ways. It’s hypocritical for them to freely train on the world’s data while restricting others from doing the same to their models.
“While the great innovation that comes from model providers having fair use rights to train models on public data is needed, I find it ironic that the status quo is to then turn around and impose restrictive terms on distillation,” Nadella writes.
Nadella is particularly concerned when model makers “reserve the right to learn from customer usage and interaction data.”
Nadella’s solution is the kind of thing the CEO of a giant cloud provider would suggest. He wants companies to “retain ownership” of their data, including prompts, feedback, etc. So he’s urging them to build their own “proprietary learning environments” on the cloud (where their data is likely already stored anyway and, conveniently, could mean Microsoft’s cloud, Azure). He also wants companies to build in what he calls “orchestration layers” — essentially, a way to easily switch between AI models from different providers rather than being locked into one. Tools like AI “gateways” that let companies do exactly this have become increasingly popular.
While Nadella never uses the words “open source” as the method for retaining ownership, this is an obvious subtext. Yet, there’s another subtext.
Large companies, many of which still have some of their own data centers in addition to using the cloud, are already moving to open source models installed on their own premises (“on-prem,” in industry jargon). Idit Levine, founder and CEO of Solo.io — which makes networking and security software that helps enterprises manage AI systems — says she’s seeing exactly this shift play out with her own customers. After experimenting with proprietary model makers, they start asking themselves: “Can I take an open source model and run it on-prem? It will do almost 90% of what the big one’s doing. It will cost way less,” she tells TechCrunch. “They understand that, and they can control it.”
Solo.io’s technology was selected last year to be the tech powering the Linux Foundation’s Agentgateway project. Her company counts enterprises like T-Mobile, ADP, and SAP as customers. She sees companies increasingly installing on-premise open source models and sees it as the next big wave in enterprise AI use.
She’s not alone. Vercel (best known as a platform for building and hosting websites, which has recently added AI model-switching tools) and OpenRouter (a company that helps developers route requests across different AI models) are both seeing a surge in traffic to open source models. In fact, open models accounted for 29% of all traffic routed through Vercel’s gateway last month.
With the CEO of Microsoft, a company that has invested in both OpenAI and Anthropic, now openly urging enterprises to be wary of using proprietary models, we’ll bet this trend continues to grow. “In consuming intelligence, you are creating intelligence. And what you create should belong to you,” Nadella writes.
When you purchase through links in our articles, we may earn a small commission. This doesn’t affect our editorial independence.
Tech
US charges Russian ‘bulletproof’ web hosts over cyberattacks that netted $62M from cybercrime victims
U.S. prosecutors have charged three Russian nationals and two web hosts with hacking, conspiracy, and money laundering over their alleged roles in hosting cyberattacks that caused tens of millions of dollars in damages to U.S. businesses.
The three Russians, Alexander Volosovik, Kirill Zatolokin, and Yulia Pankova, who reside in St. Petersburg, are accused of owning and running two web hosts, Media Land and ML.Cloud, which allegedly provided criminals and state-backed hackers with web hosting and infrastructure support for carrying out cyberattacks.
The Russians were first charged in 2024, but the indictment was unsealed this week. The U.S. Treasury previously sanctioned Media Land and ML.Cloud for allowing ransomware gangs, including LockBit, BlackSuit, and Play, to use their infrastructure. Economic sanctions bar Americans and U.S. businesses from transacting with the Russians or their companies.
Prosecutors said that hackers used the web hosts to launch distributed denial-of-service attacks designed to knock websites and services offline, launch phishing attacks, and carry out cyberattacks on critical infrastructure in the United States.
Hackers used the companies to launch attacks on dozens of U.S. businesses across more than 20 states, netting some $62 million in proceeds from cybercrime.
According to the Justice Department, by offering their services as a “bulletproof” web host, the companies deliberately aimed to shield their customers from law enforcement demands and takedowns.
The web host suspects are unlikely to be captured, given that the hackers are located in Russia and extraditions to the U.S. are rare. Russia is known to shield its citizens from overseas extradition requests, but law enforcement have previously arrested high-value suspects when they travel to countries with diplomatic agreements with the United States.
In a statement, U.S. Assistant Attorney General A. Tysen Duva said the actions of the web hosts “put the American public at risk,” adding: “We will continue to dismantle these networks and protect our critical infrastructure from cybercriminals at home and abroad.”
When you purchase through links in our articles, we may earn a small commission. This doesn’t affect our editorial independence.
Tech
India bets billions on breaking China’s grip on smartphone manufacturing
India unveiled billions of dollars in new incentives for smartphone manufacturing and an expanded semiconductor push on Wednesday, seeking to build on its success assembling Apple’s iPhones and draw more of the global electronics supply chain away from China.
Called the Mobile Phone Manufacturing Scheme, the ₹625 billion (about $6.5 billion) program will run for five years and reward smartphone manufacturers based on eligible sales, with incentives ranging from 2.25% to 5% and an additional 1.5% for sourcing key components and sub-assemblies in India. New Delhi also committed a further ₹1.28 trillion (around $13.3 billion) to bolster domestic semiconductor manufacturing, expanding a $10 billion chip incentive program launched in 2021 with greater support for chip equipment, materials, design, and research.
Over the past decade, India has emerged as an increasingly important smartphone manufacturing hub, drawing production from Apple, Samsung, and Chinese brands including Xiaomi, Oppo, and Vivo. Apple began assembling iPhones in the country in 2017 and has since expanded production through suppliers including Foxconn and India’s Tata Group, with about 25% of its iPhones now made in India as the Cupertino company diversifies its supply chain beyond China.
The manufacturing push is broadening beyond Apple. Last week, the Indian government cleared a smartphone manufacturing joint venture between China’s Vivo and Indian electronics maker Dixon Technologies. New Delhi also scrapped import duties on some phone and electronics components, a move that could lower production costs for companies, including Apple and Xiaomi.
Still, India has a long way to go before it can challenge China’s dominance. China accounted for 63% of global smartphone production in 2025, compared with India’s 18%, according to Counterpoint Research, underscoring the scale of the manufacturing and supplier ecosystem New Delhi is trying to build.
The new program marks a shift from the “assemble more” playbook that defined India’s earlier manufacturing incentives toward “depth, R&D and local value capture,” said Navkendar Singh, associate vice president at research firm IDC. India has excelled at final assembly while remaining reliant on imported components, he told TechCrunch.
“Apple stands to benefit directly,” Singh said, adding that India’s strengthening manufacturing and export credentials could give the company greater confidence to diversify production away from China, while incentivizing its supply-chain partners to source more components locally.
The smartphone manufacturing program will run through March 2031. The Indian government expects mobile-phone production during that period to total about ₹39 trillion (around $405 billion) and the scheme to create about 60,000 direct jobs.
The five-year program could help generate stronger long-term returns for India’s component ecosystem and attract more manufacturers to the country, Tarun Pathak, research director at Counterpoint Research, told TechCrunch.
Smartphone brands are looking to “save every cent” on component sourcing as memory prices reach record highs, Pathak said. Local production, he noted, could offer advantages over the longer term, particularly as a weaker Indian rupee raises the cost of imports.
In addition to incentivizing local manufacturing, New Delhi wants domestic companies to capture more of the value in the smartphone industry. The government plans to foster homegrown mobile-phone brands, Indian IT Minister Ashwini Vaishnaw said at a media briefing announcing the new manufacturing initiatives. The smartphone program includes an additional incentive of 3% of eligible sales for product design and research aimed at developing Indian brands.
India has had homegrown handset makers including Micromax, Karbonn, and Lava. However, Indian brands lost significant ground as Chinese rivals such as Xiaomi, Vivo, and Oppo expanded aggressively in the country and now account for much of the smartphone market.
The smartphone industry’s ambitions extend well beyond creating domestic brands. India should aim to account for 35% to 40% of global mobile-phone production, said Pankaj Mohindroo, chairman of the India Cellular and Electronics Association, whose members include Apple and Google.
The new policy could help build the supplier networks, engineering expertise, and manufacturing know-how needed to deepen India’s role in global supply chains, Mohindroo said.
India’s parallel bets on mobile phones and semiconductors show New Delhi is trying to build the deeper electronics manufacturing ecosystem that has underpinned China’s dominance. Its iPhone assembly boom proved the country can win a bigger role in global manufacturing. The harder test will be whether the suppliers, technology, and higher-value production follow.
When you purchase through links in our articles, we may earn a small commission. This doesn’t affect our editorial independence.
Tech
Spotify expands parent-managed accounts to users on its free tier
Spotify announced on Wednesday that it’s bringing parent-managed accounts for kids to its free tier. Families in the U.S., U.K., Australia, France, Germany, and the Netherlands can now create a “Managed Account” for their child, a feature previously available only to paid subscribers.
“Managed Accounts,” which launched in 2024, is a shared account feature that allows parents to control what their children listen to.
Because these accounts are separate, kids’ music choices won’t impact their parents’ algorithm or show up in their annual Spotify Wrapped experience. Children can add songs to their favorites, create their own playlists, and have their own personalized recommendations.
The expansion of Managed Accounts to free users reflects broader efforts by major tech companies to give parents greater control over how their children use online platforms, and which features are available to them in response to regulatory pressure.
With Managed Accounts, parents can control and restrict playback of specific artists and songs. By default, children can’t listen to music labeled as explicit, and video playback is also disabled by default. Interactivity features are also limited on managed accounts, which means that kids don’t get access to age-gated features like Messages.
Managed Accounts give parents more granular control over the music their child can listen to, without requiring them to use the more restrictive Spotify Kids app.
To set up the managed account, Family Plan account holders need to navigate to their account pages in the app, select the “Add a Member” option, and tap the “Add a listener aged under 13 (or the market equivalent)” option. From there, parents will be guided through some steps to get their child’s account set up, including choosing a display name and setting up content preferences. Parents have the option to make adjustments at any time.
Spotify says it plans to bring Managed Accounts to more countries soon.
When you purchase through links in our articles, we may earn a small commission. This doesn’t affect our editorial independence.
US charges Russian ‘bulletproof’ web hosts over cyberattacks that netted $62M from cybercrime victims
How Star Trek’s Sexiest Episode Fixed The Franchise's Biggest Missed Opportunity
