Neuralink logo displayed on a phone screen, a silhouette of a paper in shape of a human face and a binary code displayed on a screen are seen in this multiple exposure illustration photo taken in Krakow, Poland on December 10, 2021.
Jakub Porzycki | Nurphoto | Getty Images
Elon Musk’s health tech venture Neuralink shared updates to its brain-implant technology during a “show and tell” recruitment event Wednesday night. Musk said during the event that he plans to get one of the implants himself.
Musk said two of the company’s applications will aim to restore vision, even for people who were born blind, and a third application will focus on the motor cortex by restoring “full body functionality” for people with severed spinal cords. “We’re confident there are no physical limitations to restoring full body functionality,” Musk said.
Neuralink could begin to test the motor cortex technology in humans in as soon as six months, Musk said.
“Obviously, we want to be extremely careful and certain that it will work well before putting a device in a human, but we’re submitted, I think, most of our paperwork to the FDA,” he said.
But he also claimed he plans to get one himself. “You could have a Neuralink device implanted right now and you wouldn’t even know. I mean, hypothetically. In one of these demos, in fact, in one of these demos, I will,” he said. He reiterated that on Twitter after the event.
Since none of Neuralinks’ devices have been tested on humans or approved by the FDA, Wednesday’s announcements warrant skepticism, said Xing Chen, assistant professor in the Department of Ophthalmology at the University of Pittsburgh School of Medicine.
“Neuralink is a company, it doesn’t have to answer to shareholders,” she told CNBC. “I don’t know how much oversight is involved, but I think it’s very important for the public to always keep in mind that before anything has been approved by the FDA, or any governmental regulatory body, all claims need to be very, very skeptically examined.”
Neuralink was founded in 2016 by Musk and a group of other scientists and engineers. It strives to develop brain-computer interfaces, or BCIs, that connect the human brain to computers that can decipher neural signals.
Musk invested tens of millions of his own personal wealth into the company and has said, without evidence, that Neuralink’s devices could enable “superhuman cognition,” enable paralyzed people to operate smartphones or robotic limbs with their minds someday, and “solve” autism and schizophrenia.
The company’s presentation Wednesday echoed these lofty ambitions, as Musk claimed that “as miraculous as it may sound, we’re confident that it is possible to restore full body functionality to someone who has a severed spinal cord.”
Musk showed footage of a monkey with a computer chip in its skull playing “telepathic video games,” which Neuralink first debuted over a year ago. The billionaire, who is also the CEO of Tesla, SpaceX and the new owner of Twitter, said at the time that he wants to implant Neuralink chips into quadriplegics who have brain or spinal injuries so that they can “control a computer mouse, or their phone, or really any device just by thinking.”
Neuralink has come under fire for its alleged treatment of monkeys, and the Physician’s Committee for Responsible Medicine called on Musk Wednesday to release details about experiments on monkeys that had resulted in their internal bleeding, paralysis, chronic infections, seizures, declining psychological health and death.
Jeff Miller/University of Wisconsin-Madison
Neuralink’s flashy presentations are unusual for companies in the medical devices space, said Anna Wexler, an assistant professor of Medical Ethics and Health Policy at the Perelman School of Medicine at the University of Pennsylvania. She said it’s risky to encourage people who have serious disabilities to get their hopes up, especially if they could possibly incur injuries as the technology is implanted during surgery.
Wexler encouraged people to put on their “skeptic hat” about Neuralink’s big claims.
“From an ethical perspective, I think that hype is very concerning,” she said. “Space or Twitter, that’s one thing, but when you come into the medical context, the stakes are higher.”
Chen, who specializes in brain-computer interfaces, said Neuralink’s implants would require subjects to undergo a very invasive procedure. Doctors would need to create a hole in the skill in order insert the device into the brain tissue itself.
But even so, she thinks some people would be willing to take the risk.
“There’s quite a few disorders such as epilepsy, Parkinson’s and obsessive-compulsive disorder in which people have received brain implants and the disorders have been treated quite successfully, allowing them to have an improved quality of life,” Chen said. “So I do feel that there is a precedent for doing this.”
Wexler said she believes the decision would ultimately come down to an individual patient’s personal risk-benefit calculation.
Neuralink is not the only company trying to innovate using brain-computer interfaces, and many have made big strides in recent years. Blackrock Neurotech is on track to bring a BCI system to market next year, which would make it the first commercially available BCI in history. Synchron received FDA approval in 2021 to begin a clinical trial for a permanently implanted BCI, and Paradromics is reportedly gearing up to begin in-human testing in 2023.
Meta Ray-Ban Gen 2 AI glasses during the Meta Connect event in Menlo Park, California, US, on Wednesday, Sept. 17, 2025.
David Paul Morris | Bloomberg | Getty Images
EssilorLuxottica said a healthy amount of its revenue growth in the third quarter was due to its partnership with Meta, primarily from its Ray-Ban brand, to develop and sell smart glasses.
“Clearly there is a lift coming from Ray-Ban Meta wearables as a product category,” CFO Stefano Grassi said on the company’s third-quarter earnings call.
The European eyewear company said sales in in the quarter grew 11.7% year-over-year to 6.9 billion euros (about $8 billion) from 6.44 billion euros a year earlier. Of that growth, more than 4 percentage points came from wearables, which includes the Meta products, the company said.
In 2019, Meta and Luxottica inked a deal for Ray-Ban Meta branded smart glasses. Most recently, Luxottica’s Oakley brand has joined the partnership, with the debut in June of the Oakley Meta HSTN smart glasses. The companies are also working on a version of the smart glasses to be released under the Prada brand, CNBC reported in June.
Luxottica, which also oversees several popular brands like Vogue Eyewear and Persol, has been heavily pushing internet-connected glasses that work with Meta’s AI-powered digital assistant. The technology allows users to play music, take photos and perform other actions similar to how they would use smartphones.
“We believe that glasses will be the future,” Grassi said, adding that the wearables business is profitable. “Glasses will materially replace most of the functionality that today we have embedded into our phones.”
Grassi’s statement echoes sentiments expressed by Meta CEO Mark Zuckerberg, who said in July that “Personal devices like glasses that understand our context because they can see what we see, hear what we hear, and interact with us throughout the day will become our primary computing devices.”
A couple weeks into the fourth quarter, Grassi said he has “a good degree of optimism” for the period, in part because of the rollout of “all the new products that have been recently presented at the Meta Connect,” which will “all play a role in our fourth-quarter profile.”
At the Connect event in September, Zuckerberg revealed the $799 Meta Ray-Ban Display glasses, which have a small digital display that can be manipulated with an accompanying wristband powered by neural technology.
The company also unveiled new smart glasses, including the $499 Oakley Meta Vanguard glasses and the $379 Ray-Ban Meta (Gen 2) glasses.
Grassi said that Luxottica’s sales growth in North America in the third quarter had more to do with the Ray-Ban Meta glasses than the effects of tariffs, which led to higher prices for its products.
He said the company will be able to reach the 10 million unit capacity that it had originally planned to hit by the end of 2026 earlier than anticipated.
“The overall ecosystem of wearables is going to bring not only revenue associated with the hardware but also the revenue associated with lenses” and over time from services tied to AI.
EssilorLuxottica shares rose 2.4% on Thursday.
Meta isn’t the only tech giant getting into the burgeoning smart glasses market.
Alphabet announced in May a $150 million partnership with Warby Parker to develop smart glasses powered by Google’s Gemini AI digital assistant, while China’s Alibaba unveiled its smart glasses in July that utilize its Quark AI assistant. Apple and OpenAI are also reportedly developing smart glasses.
Oracle CEO Clay Magouyrk, center, speaks on a media tour of the Stargate data center in Abilene, Texas, on Sept. 23, 2025. Stargate is a collaboration of OpenAI, Oracle and SoftBank, with promotional support from President Donald Trump, to build data centers and other infrastructure for artificial intelligence throughout the US.
Kyle Grillot | Bloomberg | Getty Images
Oracle shares ended Thursday trading up 3% as it called for more business in core categories and confirmed a cloud-computing deal with social media company Meta.
The maker of database software sees $20 billion in artificial intelligence-powered database and AI data platform revenue in the 2030 fiscal year, up from $2.4 billion in fiscal 2025 and $3 billion in fiscal 2026.
“You see the change in these numbers that it’s a little bit easier for us to find supply, not this year or next year, but in subsequent years,” Clay Magouyrk, one of Oracle’s two new CEOs, told analysts Thursday at the company’s AI World conference in Las Vegas. “So as we’re able to find that supply, customers contract for it, we see immense demand, and then we go about delivering that to customers.”
Magouyrk said that in 30 days during the current quarter, Oracle contracted $65 billion in new cloud infrastructure commitments.
“It was across seven different contracts from four different customers,” Magouyrk said. “None of those customers are OpenAI. I know some people are questioning sometimes, ‘Hey, is it just OpenAI? The reality is, we think OpenAI is a great customer, but we have many customers.”
Meta which operates Facebook and Insatgram is one of the four customers, he said. Bloomberg reported in September that the two companies were discussing a $20 billion deal.
The deal with Meta comes amid a flurry of spending by tech companies to invest in the infrastructure for their AI initiatives. Meta in July said that it expects to spend between $66 billion and $72 billion this year in capital expenditures.
In recent years, Oracle has expanded its cloud infrastructure division that competes with the likes of Amazon and Google. At the same time, Oracle has started offering its database in clouds other than its own.
Oracle secured a commitment from OpenAI in excess of $300 billion in July.
AI infrastructure has an adjusted gross margin of 30% to 40% after land, data center, power and computing equipment costs, Oracle said. Earlier this month, The Information reported that Oracle saw a 14% gross margin on renting out Nvidia AI chips in the August quarter.
“I’ve read a lot of stories that are speculating that Oracle is chasing revenue for revenue’s sake, but let’s be crystal clear,” said Doug Kehring, the company’s principal financial officer. “We only pursue opportunities where we have a clear line of sight to attractive market margins that reward us for intellectual property and the activity we bring to customers.”
After market close, Oracle said it’s now targeting $21 in adjusted earnings per share on $225 billion in revenue for fiscal 2030, representing a 31% compound annual growth rate. Analysts polled by LSEG were looking for $18.92 per share on $198.39 billion in revenue. The stock slipped 2% in extended trading.
U.S. cybersecurity company F5 fell 12% on Thursday after disclosing a system breach in which a “highly sophisticated nation-state threat actor” gained long-term access to some systems.
F5 shares were pacing for the worst day since April 27, 2022, when the stock fell 12.8%.
The company disclosed the breach in a Securities and Exchange Commission filing on Wednesday and said the hack affected its BIG-IP product development environment. F5 said the attacker infiltrated files containing some source code and information on “undisclosed vulnerabilities” in BIG-IP.
The breach was later attributed to state-backed hackers from China, Bloomberg reported, citing people familiar with the matter.
F5, which was made aware of the attack in August, said they have not seen evidence of any new unauthorized activity.
Read more CNBC tech news
“We have no knowledge of undisclosed critical or remote code vulnerabilities, and we are not aware of active exploitation of any undisclosed F5 vulnerabilities,” F5 said in a statement.
The cybersecurity giant told customers that hackers were in the network for at least 12 months and that the breach used a malware called Brickstorm, according to Bloomberg.
F5 would not confirm the information.
Brickstorm is attributed to a suspected China-nexus threat dubbed UNC5221, Google Threat Intelligence Group said in a blog post. The malware is used for maintaining “long-term stealthy access” and can remain undetected in victim systems for an average of 393 days, according to Mandiant.
The attack prompted an emergency directive from the Cybersecurity and Infrastructure Security Agency on Wednesday, telling all agencies using F5 software or products to apply the latest update.
“The alarming ease with which these vulnerabilities can be exploited by malicious actors demands immediate and decisive action from all federal agencies,” CISA Acting Director Madhu Gottumukkala said. “These same risks extend to any organization using this technology, potentially leading to a catastrophic compromise of critical information systems.”
The UK’s National Cyber Security Centre also issued guidance for the F5 attack, advising customers to install security updates and continue monitoring for threats.