The Twitter profile page belonging to Elon Musk is seen on an Apple iPhone mobile phone.
Nurphoto | Nurphoto | Getty Images
New Twitter owner Elon Musk has pulled more than 50 of his trusted Tesla employees, mostly software engineers from the Autopilot team, into his Twitter takeover, CNBC has learned.
Musk, who is CEO of automaker Tesla and reusable rocket maker SpaceX, completed the $44 billion acquisition of Twitter on Oct. 28 and made his mark there immediately. He fired the company’s CEO, chief financial officer, policy and legal team leaders right away, and has also dissolved Twitter’s board of directors.
According to internal records viewed by CNBC, employees from Musk’s other companies are now authorized to work at Twitter, including more than 50 from Tesla, two from the Boring Company (which is building underground tunnels) and one from Neuralink (which is developing a brain-computer interface).
Some of Musk’s friends, advisors and backers, including the head of his family office Jared Birchall, angel investor Jason Calacanis, and founding PayPal chief operating officer and venture capitalist David Sacks, are also involved. So are two people who share Musk’s last name, James and Andrew Musk, who have worked at Palantir and Neuralink, respectively.
Among the dozens who Elon Musk enlisted specifically from Tesla are: director of software development Ashok Elluswamy, director of Autopilot and TeslaBot engineering Milan Kovac, senior director of software engineering Maha Virduhagiri; Pete Scheutzow, a senior staff technical program manager, and Jake Nocon, who is part of Tesla’s surveillance unit, as a senior manager of security intelligence.
Nocon previously worked for Uber and Nisos, a security company that had a multimillion-dollar contract with Tesla to identify insider threats, and monitor critics of the company.
At Twitter, Musk is counting on his lieutenants and loyalists to decide who and what to cut or keep at the social network.
He is also pressing them to learn everything they can about Twitter as quickly as possible, from source code to content moderation and data-privacy requirements, so he can redesign the platform, several Twitter employees told CNBC over the weekend.
Musk has billed himself as a free speech absolutist, but he has to balance those wishes with laws and business realities. He said in an open letter to advertisers last week as he was taking over the company: “Twitter obviously cannot become a free-for-all hellscape, where anything can be said with no consequences.”
It is not immediately clear how Tesla employees are expected to split their schedules between the automaker and Twitter.
Typically, when Tesla employees work for other Elon Musk ventures, usually SpaceX or the Boring Company, they can get paid by the other venture as a consultant. Some of Musk’s employees have full-time roles at more than one of his businesses. For example, Tesla Vice President of Materials Charlie Kuehmann, is concurrently a vice president at SpaceX.
At other times, two Tesla employees told CNBC, workers at the electric automaker are pressured to help with projects at his other companies for no additional pay because it’s seen as good for their careers, or because the work is regarded as helping with a related party transaction or project.
Tesla is facing serious scrutiny around the technology built and maintained by its Autopilot team, namely its driver-assistance systems, which are marketed as Autopilot, FSD and FSD Beta.
The Securities and Exchange Commission, the Department of Justice, and the California Department of Motor Vehicles are all investigating whether Tesla or Musk violated laws and misled consumers about Tesla’s driver assistance systems, which are still in development and do not make the company’s cars self-driving.
Meanwhile, the federal vehicle safety authority, NHTSA, continues to investigate whether Tesla driver assistance systems may have contained defects that contributed to or caused collisions. The way that Tesla marketed these systems on social media, including Twitter, is part of the scope of at least one NHTSA investigation.
Code reviews and 12-hour shifts
Several Twitter employees told CNBC over the weekend that Tesla employees now at Twitter have been involved in code review at the social network, even though their skills from working on Autopilot and other Tesla software and hardware do not directly overlap with the languages and systems used to build and maintain the social network. These employees asked not to be named because they’re not authorized to talk to the press about internal matters, and feared retaliation.
For example, most engineers in automotive companies, even the tech-forward Tesla, do not have experience designing and operating search engines and platforms that are broadly accessible to the public.
Twitter has multiple code bases with millions of lines of code in each, and myriad 10- or even 100-query per second (QPS) systems underpinning it. At Tesla, Python is one of the preferred scripting languages, and at Twitter programmers have used Scala extensively.
Twitter also has more exposure to international regulations around hate speech and data privacy, for example, particularly the European Union’s General Data Protection Regulation.
Twitter employees who were there before Musk took over said they have been asked to show his teams all manner of technical documentation, to justify their work and their teams’ work, and to explain their value within the company. The threat of dismissal looms if they do not impress, they said.
The employees said they are worried about being fired without cause or warning, rather than laid off with severance. Some are worried that they will not be able to reap the rewards of stock options that are scheduled to vest in the first week of November, according to documentation viewed by CNBC.
Meanwhile, the Twitter employees said they have not received specific plans from Musk and his team yet, and are largely in the dark about possible head count cuts within their groups, budgets and long-term strategies.
Musk has set nearly impossible deadlines for some to do-list items, however.
One immediate project is to redesign the company’s subscription software, dubbed Twitter Blue, and the company’s verification system (known sometimes as “blue checks” for the way they are denoted on the service). Employees say Musk wants that work done by the first week of November. The Verge previously reported that Musk wants to charge $20 per user per month, and to only give verification marks to the accounts of users who are paid subscribers, and would remove verification from accounts who do not pay for Twitter Blue.
Managers at Twitter have instructed some employees to work 12-hour shifts, seven days a week, in order to hit Musk’s aggressive deadlines, according to internal communications. The sprint orders have come without any discussion about overtime pay or comp time, or about job security. Task completion by the early November deadline is seen as a make-or-break matter for their careers at Twitter.
In an atmosphere of fear and distrust, many Twitter employees have stopped communicating with each other on internal systems about workplace issues. What’s more, some of Twitter’s Slack channels have gone nearly silent, multiple employees told CNBC.
Meanwhile, Musk and his inner circle have been plumbing archived messages in the systems, ostensibly looking for people to fire and budgets or projects to slash.
On Sunday night, in a display of his unfettered access to internal information at the company, CEO Elon Musk (who calls himself “Chief Twit” but is officially CEO and sole director) posted a screenshot to his 112 million listed followers on Twitter.
The screenshot depicted comments made by Twitter’s head of safety and integrity, Yoel Roth, in May 2022. At the time, Musk was trying to get out of his agreement to buy Twitter for $54.20 per share.
In court, and in public, Musk had vociferously accused Twitter of faking metrics, specifically of playing down the amount of spam, fake accounts and harmful bots that exist on the platform.
In the internal message that Musk made public, Roth wrote disparagingly of a person involved with the business named Amir, and also remarked, that if Amir continued to “BS” him or others about objectives and key results, Twitter would be “literally doing what Elon is accusing us of doing.”
Musk alleged in a tweet that, “Wachtell & Twitter board deliberately hid this evidence from the court.” He also appeared to threaten further legal action, writing: “Stay tuned, more to come…”
Representatives for Twitter, Tesla and the law firm Wachtell, Lipton, Rosen & Katz have yet to respond to requests for comment.
Participants at the presentation of new iPhone models from Apple try out the new thinner iPhone Air.
Andrej Sokolow | Picture Alliance | Getty Images
Apple has postponed the launch of its new iPhone Air model in China due to regulatory issues surrounding its eSIM design, the company said.
Wireless carriers in China need a special license from the government before they can sell a new device with an eSIM, and the carriers haven’t secured that approval yet, Apple said. The company added that it’s working to make the device available in China as soon as possible.
Apple announced the iPhone Air at its annual event on Tuesday. The device, which is 5.6 millimeters thick, marks the first major new iPhone design since the iPhone X was introduced in 2017. The iPhone Air doesn’t support a physical SIM card, and instead features an eSIM built into the device.
CEO Tim Cook told CNBC’s Jim Cramer on Friday that the eSIM is what allows the device to still have “great” battery life.
“It’s eSIM only, and so we were able to take the battery and extend the battery to areas that previously had the physical cell,” Cook said.
Read more CNBC tech news
The company previously said the iPhone Air would become available for pre-order in the region on Friday at 2 a.m. EST before it goes on sale September 19.
As of Friday morning, the iPhone Air product page on Apple’s China website stated, “Release information will be updated later.”
The website notes that China Mobile, China Telecom and China Unicom will offer eSIM support for the iPhone Air, “with specific timing subject to regulatory approval.”
Jensen Huang, co-founder and chief executive officer of Nvidia, at the London Tech Week exposition in London, UK, on Monday, June 9, 2025.
Bloomberg | Bloomberg | Getty Images
Nvidia and OpenAI are in discussions about backing a major investment in Britain focused on boosting artificial intelligence infrastructure in the country.
The two tech firms are discussing a sizable deal to support data center development in the country which could ultimately be worth billions of dollars, a person familiar with the matter told CNBC, confirming earlier reporting from the Financial Times.
The companies are still working through various processes at the moment with Nvidia and cloud computing firm Nscale, said the person, who did not want to be named due to the sensitivity of the issue.
They added that an investment agreement has not yet been finalized. It is expected to be unveiled next week during U.S. President Donald Trump’s state visit to the U.K.
Nvidia and Nscale did did not immediately respond to CNBC’s request for comment. OpenAI declined to comment on the discussions.
Countries around the world have been courting major U.S. AI players in a bid to boost their own national infrastructure and technological ambitions.
The topic of so-called “sovereign” AI — the idea of onshoring the data processing infrastructure behind advanced artificial intelligence systems — has been top of mind for officials as governments look to reduce their dependency on foreign countries for critical technologies.
The U.K. government declined to comment when asked by CNBC about the investment discussions with OpenAI, Nvidia and Nscale. Nvidia CEO Jensen Huang is set to join Trump on his state visit to Britain next week.
Earlier this year, the Nvidia boss called the U.K. an “incredible place to invest” and said his multitrillion-dollar chipmaker would boost investment in the country. “The U.K. is in a Goldilocks circumstance,” Huang said at the time in a panel discussion with British Prime Minister Keir Starmer.
Apple AirPods Pro 3 models are displayed during Apple’s “Awe-Dropping” event at the Steve Jobs Theater on the Apple Park campus in Cupertino, California, on Sept. 9, 2025.
Nic Coury | AFP | Getty Images
For decades, shows like “Star Trek” and novels like “The Hitchhikers Guide to the Galaxy” have showcased fictional universal translators, capable of seamlessly converting any language into English and vice versa.
Now, those gadgets once limited to works of science fiction are inching close to reality.
During its iPhone unveiling event on Tuesday, Apple included a video of many travelers’ dream scenario. It showed an English-speaking tourist buying flowers in an unnamed Spanish-speaking country. The florist addressed the tourist in Spanish, but what the tourist heard was in clear, coherent English.
“Today all the red carnations are 50% off,” the tourist heard in English in her headphones, at essentially the same time that the clerk was speaking.
The video was marketing material for Apple’s latest AirPods Pro 3, but the feature is one of many of its kind coming from tech companies that also include Google parent Alphabet and Meta, which makes Facebook and Instagram.
Apple introduces live translation to airpods.
Courtesy: Apple
Technological advancements spurred by the arrival of OpenAI’s ChatGPT in late 2022 have ushered in an era of generative artificial intelligence. Almost three years later, those advancements are resulting in real-time language translators.
For Apple, Live Translation is a key selling point for the AirPods Pro 3, which the company unveiled on Tuesday. The new $250 earbuds go on sale next week, and with Live Translation, users will be able to immediately hear French, German, Portuguese and Spanish translated to English. Live Translation will also arrive as an update to AirPods 4 and AirPods Pro 2 on Monday.
And when two people are speaking to each other wearing AirPods, the conversation can be translated both ways simultaneously inside each user’s headphones. In Apple’s video demo, it looked like two people talking to each other in different languages.
Analysts are excited that the feature could mark a step forward for Apple’s AI strategy. The translation feature needs to be paired with a new-enough iPhone to run Apple Intelligence, Apple’s AI software suite.
“If we can actually use the AirPods for live translations, that’s a feature that would actually get people to upgrade,” DA Davidson analyst Gil Luria told CNBC on Wednesday.
Translation is emerging as a key battleground in the technology industry as AI gets good enough to translate languages as quickly as people speak.
But Apple is not alone.
Host Jimmy Fallon holds Pixel 10 Pro Fold mobile phone during the ‘Made by Google’ event, organised to introduce the latest additions to Google’s Pixel portfolio of devices, in Brooklyn, New York, U.S., August 20, 2025.
Brendan McDermid | Reuters
A crowded market
In the past year, Google and Meta have also released hardware products featuring real-time translation capabilities.
Google’s Pixel 10 phone has a capability that can translate what a speaker is saying to the listener’s language during phone calls. That feature, called Voice Translate is designed to also preserve the speaker’s voice inflections. Voice Translate will start showing up on people’s phones through a software update on Monday.
In Google’s live demo in August, Voice Translate was able to translate a sentence from entertainer Jimmy Fallon into Spanish, and it actually sounded like the comedian. Apple’s feature does not try to imitate the user’s voice.
Meanwhile, Meta in May announced that its Ray-Ban Meta glasses would be able to translate what a person is saying in another language using the device’s speakers, and the other party in the conversations would be able to see translated responses transcribed on the user’s phone.
Meta will hold its own product keynote on Wednesday, where the company is expected to announce the next generation of its smart glasses, which will feature a small display in one of the lenses, CNBC reported in August. It’s unclear if Meta will announce more translation features.
Meta employee Sara Nicholson poses with the Ray-Ban sunglasses at the Meta Connect annual event at the company’s headquarters in Menlo Park, California, U.S., September 24, 2024.
Manuel Orbegozo | Reuters
And OpenAI in June showcased an intelligent voice assistant mode for ChatGPT that has fluid translation built-in as one of many features. ChatGPT is integrated with Apple’s Siri, but not in voice mode. OpenAI is planning to release new hardware products with Apple’s former design guru Jony Ive in the coming years.
The rise of live translation could also reshape entire industries. Translators and interpreters are the number one type of job threatened by AI, and 98% of translators’ work activities overlap with what AI can do, a Microsoft Research study published in August found.
Purpose built translators
In the past several years, a number of purpose-built translation gadgets have entered the market, taking advantage of global high-speed cellular service and improving online translation services to produce puck-like devices or headphones with translation built-in for a couple hundred dollars.
“What I love about what Apple is doing is it really just illuminates the fact that how pressing of an issue this is,” said Joe Miller, U.S. general manager of Japan-based Pocketalk, which makes a $299 translation device that goes between two people conversing in different languages and translates their conversation in audio and text.
Given Apple’s massive scale and the fact that the Apple shipped about 18 million sets of wireless headphones in the first quarter alone, according to Canalys, the company’s entry into the market will expose a wider subset of customers to improvements translation tech has made in recent years.
Despite Apple’s entry into the market, makers of purpose-built devices say their focus on accuracy and knowledge of linguistics will provide better translations than what’s available for free with a new phone.
“We actually hired linguists,” said Aleksander Alski, head of U.S. and Canada for Poland-based Vasco Electronics, which is releasing translation headphones that can imitate the user’s voice, like Google’s feature. “We combined the AI with with human input, and thanks to that, we were able to secure much higher accuracy throughout all the languages we offer.”
There’s also home-field advantage. Vasco Electronics’ largest market is Europe, and Apple’s Live Translation isn’t available for EU users, Apple said on its website.
Some of the products being introduced by tech companies are less than universal, and are limited to a small number of languages for now. Apple’s feature is only available in 5 languages, versus Pocketalk’s 95.
Pocketalk’s Miller believes that the potential of the technology goes far beyond a tourist ordering a glass of wine in France. He says that it’s most powerful when its used in workplaces like schools and hospitals, which require privacy and security features that go beyond what Apple and Google provide.
“This isn’t about luxury tourism and travel,” Miller said. “This is about the intersection of language and friction, when a discussion needs to be had.”