It’s been looming for weeks, but now the end is near: Just a few hundred Tesla Model S and Model X vehicles remain unsold. Tesla CEO Elon Musk confirmed this week in a post on X that custom orders of the Model S sedan and Model X SUV are over. “All that’s left are some in inventory,” he wrote.

Musk first announced Tesla’s plan to end Model S and Model X production back in January. And the data helps explain why.

Sales of the Tesla Model X and Model S have fallen steadily over the years as the company’s high volume and cheaper entries — the Model 3 and Model Y — took over. Tesla doesn’t separate S and X sales, instead combining them under “other models,” a category that now includes the Cybertruck. And those combined figures show S and X sales peaking in 2017 at 101,312 vehicles before declining to 50,850 vehicles (including Cybertruck) in 2025 — a fraction of the 1.63 million vehicles it delivered globally last year.

In other words, their deaths were inevitable. What comes next is a bit more complicated.

Musk isn’t filling the void left by the Model X and Model S with a traditional EV; he ditched plans to produce a lower-cost EV that was expected to be priced around $25,000. Instead, Musk is placing his bets on the Optimus robot, which has yet to go into production, and the Cybercab, an all-electric two-seater autonomous vehicle that was first shown as a concept in 2024.

Tesla plans to build Optimus robots at its Fremont, California, factory once production of the Model S and Model X end, which could be any day now that final orders have been taken. Musk has said Tesla will begin producing the Cybercab this month at its factory in Austin, Texas. 

A look back

The Model S and X EVs have taken a backseat to the more affordable Model 3 and Model Y vehicles. But their debuts, and initial sales, marked two critical moments in Tesla’s colorful and often volatile history. The Model S launched in 2012 as its first volume EV. Its popularity not only changed how consumers viewed EVs, it prompted legacy automakers — long dismissive of the value of electric vehicles — to take notice.

The Model X followed in fall 2015 and was famously described by Musk as the Fabergé egg of EVs.

“I think we got more carried away with the X,” Musk said in a September 2015 press interview attended by this reporter just an hour before Tesla’s Model X delivery event began. “I’m not sure anyone should make this car.”

The Model X was often delayed, and initially criticized for its complexity. But it ultimately introduced the company to a new market: women.

The Model X raised Tesla’s profile, and it set the company up for its next big move: an affordable mass-produced EV. The Model 3 had a difficult start, but it ended up catapulting Tesla into the mainstream. The Model Y clinched its status, helping Tesla widen the gap as the top-selling EV producer globally until China’s BYD took over that top global EV sales spot in 2025 when it delivered 2.26 million EVs.

Tesla continues to sell thousands of Model 3 and Model Y, but its growth has stalled, and even reversed. The company reported in January that it sold 1.69 million vehicles in 2025, a decrease for the second year in a row. Its efforts to boost sales with cheaper, stripped-down versions of the Model 3 and Model Y that were introduced in October have had a modicum of success, according to first-quarter 2026 figures that were reported April 2.

Tesla delivered 358,023 EVs globally in the first three months of the year, about 6% more than the same period in 2025, which also happened to be the company’s worst quarter in years. The figure was below analysts’ expectations of around 368,000.

But never mind that. In Musk’s view — one which he is well compensated for — Tesla isn’t an automaker or a sustainable energy company, as he has described it before. Tesla is an AI company and his new gambit goes all in on that mission.

Cybercab risks

The Optimus robot is one part of the Tesla AI effort. But it’s perhaps the Cybercab that best embodies, and exposes the risks of, the company’s AI-first campaign.

The Cybercab was designed to be used as an autonomous vehicle without traditional controls like a steering wheel or pedals — meaning once it launches it will be without the initial backup of a human safety operator.

The first Cybercab rolled off the Tesla factory assembly line in February and is supposed to go into mass production this month. Although that date could slip, as so many have in Tesla’s history.

Unlike Tesla’s previous vehicles, the challenges aren’t in its production (who can forget the production hell of the Model 3). Instead, it faces a major regulatory hurdle before it can ever hit the road. Federal motor vehicle safety standards place requirements on vehicles such as having a steering wheel and pedals. There is no evidence that Tesla has applied for an exemption, according to publicly available files with the Federal Register and the National Highway Traffic Safety Administration.

The vehicles will also rely on Tesla’s Full Self-Driving software to navigate public streets and safely shuttle passengers to their destination. Despite improvements to FSD and limited driverless robotaxi tests in Austin, Tesla has not yet demonstrated that its software can operate reliably at scale.

And that piece requires more than technical mastery. Robotaxi operations are also tricky. And in states like California, they also require permits to deploy and charge for rides in driverless vehicles.

Zoox, the autonomous vehicle company owned by Jeff Bezos’ Amazon, may end up clearing a path for Tesla and its Cybercab. Zoox received an exemption from the National Highway Traffic Safety Administration that allows the company to demonstrate its custom-built robotaxis, which lack pedals or a steering wheel, on public roads. Zoox is now going through a public process to have that exemption extended to commercial operations.

Musk tried to sell shareholders on why the risk was worth it during the company’s earnings call in January.

“The vast majority of miles traveled will be autonomous in the future,” Musk said at the time, later noting that the Cybercab is super optimized for minimum cost per mile and also for a much higher-duty cycle. “I would say probably less than, I’m just guessing, but probably less than 5% of miles driven will be where somebody’s actually driving the car themselves in the future, maybe as low as 1%.”

The Trump administration’s labor board has ordered Amazon to recognize and bargain with the International Brotherhood of Teamsters union, which represents workers at a warehouse in Staten Island. This is just the latest chapter in a multiyear standoff between Staten Island warehouse workers and Amazon, .

The union has been trying to bring Amazon to the bargaining table for years to negotiate pay, benefits and workplace safety. The labor board’s proclamation doesn’t mean that the battle is over. It’s highly likely this will be settled in court.

An Amazon spokesperson maintains that the vote to create the union was “wrong on the facts of the law” and that representatives from the National Labor Relations Board “improperly influenced the election.” The company recently stated it is “confident an unbiased court will overturn the original certification.”

Despite the eventual outcome, Teamsters President Sean O’Brien is lauding the Staten Island workers for becoming “the first group ever to force the company to recognize their union.” Workers at the facility and this was the first union victory for Amazon employees in the US.

It was considered a milestone victory for US workers across the board, given that Amazon is the . That was four years ago and led to a contracted legal battle, as Amazon has refused to recognize the union. Since that original vote, the labor board has repeatedly found that Amazon violated workers’ union rights at the Staten Island warehouse. For instance, the company didn’t pay employees when they were forced to stop working at the tail-end of 2022 and suspended 50 employees for staging a walkout due to unsafe work conditions.

There were also several harrowing incidents leading up to the union vote. It’s been reported that the company illegally fired during the . The NY Attorney General also at the warehouse to be “inadequate.” A recent study , calling out the Staten Island warehouse for dangerous working conditions. The report says that there are 7.2 serious injuries for every 100 workers.

Other US-based Amazon warehouses have yet to follow suit and unionize like Staten Island, but the same isn’t true in Canada. Workers at a warehouse in Quebec in 2024.

Anthropic has acquired Coefficient Bio, a stealth biotech AI startup founded barely eight months ago, in an all-stock deal worth just over $400 million. The acquisition brings a team of fewer than 10 people, nearly all former Genentech computational biology researchers, into Anthropic’s healthcare and life sciences division, and it signals something larger than a talent grab: the maker of Claude is staking real capital on the idea that general-purpose AI can accelerate drug discovery.

The deal, first reported by The Information on Thursday, values a company that had no publicly known product, no disclosed revenue, and no conventional traction metrics. What it did have was a founding team with rare credentials. Samuel Stanton and Nathan C. Frey, Coefficient Bio’s co-founders, both came from Prescient Design, Genentech’s computational drug discovery unit, where Frey led a multidisciplinary group working on biological foundation models and novel machine learning approaches to biomolecule design. Frey’s publication record spans more than 20 papers in journals including Science Advances and Nature Machine Intelligence, and he won an ICLR Outstanding Paper Award in 2024 for work on generative modelling for drug candidate discovery.

The startup’s stated ambition was nothing modest: artificial superintelligence for science. In practice, Coefficient Bio had built a platform enabling AI to draft drug research and development plans, manage clinical regulatory strategies, and identify new drug candidates. It was, by all accounts, a research-heavy operation that never left stealth mode.

Dimension, the New York-based venture firm founded in 2023 by former Lux Capital and Obvious Ventures partners Adam Goulburn, Zavain Dar, and Nan Li, held roughly half the company. The firm, which focuses on companies at the intersection of technology and life sciences, is now reporting a 38,513 per cent internal rate of return on the investment, a figure that says less about Coefficient Bio’s commercial viability than about the speed at which AI valuations are repricing early-stage science bets. Against Anthropic’s $380 billion post-money valuation, set in its $30 billion Series G round in February, the acquisition represents roughly 0.1 per cent dilution.

The 💜 of EU tech

The latest rumblings from the EU tech scene, a story from our wise ol’ founder Boris, and some questionable AI art. It’s free, every week, in your inbox. Sign up now!

The Coefficient Bio team will join Anthropic’s Health Care Life Sciences group, led by Eric Kauderer-Abrams, who was hired in 2025 with an explicit mandate to make Claude the dominant AI model in biology. “We want a meaningful percentage of all of the life science work in the world to run on Claude, in the same way that that happens today with coding,” Kauderer-Abrams told CNBC when Anthropic launched Claude for Life Sciences in October 2025. That platform, which integrates with tools including Benchling, PubMed, and 10x Genomics, was designed to assist researchers across the entire drug discovery pipeline, from literature review and hypothesis generation to data analysis and regulatory submissions.

The acquisition deepens that push considerably. Where Claude for Life Sciences offered a generalised research assistant, Coefficient Bio’s team brings the kind of domain-specific expertise, particularly in protein design and biomolecule modelling, that could help Anthropic build specialised tools for pharmaceutical companies willing to pay enterprise prices for AI that understands their workflows at a molecular level.

Anthropic is not entering a vacuum. Google DeepMind spun off Isomorphic Labs to pursue AI-designed drug candidates now entering clinical trials, and Nvidia announced a five-year, $1 billion partnership with Eli Lilly in January to build an AI co-innovation lab for accelerated drug discovery. OpenAI, meanwhile, has been working with Moderna to speed the development of personalised cancer vaccines. The competitive logic is straightforward: whichever foundation model becomes embedded in biopharma R&D workflows will capture an enormous and recurring revenue stream in a market where a single approved drug can generate billions.

The venture capital appetite for AI-biology crossovers is reflecting this calculus. Breakout Ventures closed a $114 million fund in March explicitly targeting early-stage biotechs that treat AI and biology as inseparable. Dimension itself is reportedly raising a $700 million third fund to double down on the same thesis. The investor conviction is that the agentic AI wave will hit life sciences as forcefully as it has hit software engineering, and the acqui-hire economics of deals like Coefficient Bio suggest the large model builders agree.

For Anthropic, the strategic arithmetic is clear enough. The company’s run-rate revenue has reached $14 billion, growing more than tenfold annually for three consecutive years, and the customer base spending over $100,000 a year on Claude has grown sevenfold. But that growth is overwhelmingly concentrated in coding, enterprise search, and general productivity. Healthcare and life sciences represent a vast adjacent market where Anthropic has laid the groundwork with Claude for Life Sciences but has not yet achieved the kind of deep integration that generates sticky, high-margin revenue.

Paying $400 million in stock for a pre-revenue team of fewer than 10 people will, understandably, invite scepticism. The price looks less like a valuation of what Coefficient Bio had built and more like a statement about what Anthropic believes it can build with the right researchers on the payroll. Whether that bet pays off depends on something the current frenzy of AI startup valuations has not yet been forced to answer: whether frontier AI models can generate genuine scientific breakthroughs, or whether they will remain very expensive literature review assistants that happen to speak the language of molecular biology.

Video making has long required breaking the bank or devoting hours to learning complex software. Google Vids eliminates all of those hassles at once. If you already have a Google account, you can create, edit, and share videos right from your browser, no downloads or subscriptions required.

Things start when you enter a brief description of what you wan to show. Google Vids creates a storyboard for you, with sequences arranged in order and ready to go. You can add any photographs or documents from your Google Drive to fill out the scenes, or let the service recommend some appropriate stock video. Then it’s just a matter of dragging clips about on a simple timeline to get the timing correct, removing any unnecessary sections, and layering in text to appear at the appropriate time.

Sale

Google Pixel 10 – Unlocked Android Smartphone – Gemini AI Assistant – Advanced Triple Rear Camera…

• Google Pixel 10 is the everyday phone unlike anything else; it has Google Tensor G5, Pixel’s most powerful chip, an incredible camera, and advanced…
• Unlocked Android phone gives you the flexibility to change carriers and choose your own data plan[2]; it works – Google Fi, Verizon, T-Mobile, AT&T…
• The upgraded triple rear camera system has a new 5x telephoto lens – up to 20x Super Res Zoom for stunning detail from far away; Night Sight takes…

Audio work is also rather simple, since you can just record a voiceover using a built-in teleprompter that scrolls along at your desired speed. The editor even has your back when it comes to filler words, detecting them automatically and allowing you to remove them with a single click. Soundtracks and sound effects are simply dropped in from a collection of alternatives that match the mood of each scenario. New video clips can be created directly from a prompt entered into the editor. Veo 3.1 creates eight-second segments with motion and native music; you receive ten of them per month for free, but higher-tier members get far more, up to a thousand per month.

On top of that, avatars add a really personal touch. You choose a character or design one from yourself, and then simply spell out the instructions, such as “take up the product and demonstrate how it works while smiling at the camera.” The avatar then follows those directions, maintaining the same face and attire throughout each shot, and interacting naturally with any accessories or backgrounds you’ve selected. You can alter the entire thing using plain text, so the same character can appear in training videos, product demos, and team updates without ever having to face a real camera.

A new Chrome addon makes screen recording incredibly simple. With one click, whatever is on your desktop, including audio, is recorded and dropped directly into the editor for chopping and assembling. Finished videos export directly to YouTube as private drafts, allowing you to review them before sharing them with the world. Google announced all of these enhancements on April 2, 2026, adding on capabilities previously launched for enterprise users and offering the service accessible to anybody who wants to turn a concept into a shareable film in minutes rather than days.

The European Union’s cybersecurity agency said Thursday that a recent hack and data breach at the EU’s executive body was the work of a cybercriminal group known as TeamPCP. 

In a new report, CERT-EU also reported that the hackers stole around 92 gigabytes of compressed data from a compromised Amazon Web Services (AWS) account used by the bloc’s executive, the European Commission, which included personal data containing names, email addresses, and the contents of emails. 

The breach affected the cloud infrastructure of the Commission’s Europa.eu platform, which member states use to host websites and publications of the bloc’s institutions and agencies.

CERT-EU wrote that the data of at least 29 other EU entities may be affected, and that dozens of internal European Commission clients could have had data stolen as well. 

The stolen data was then posted online by another hacking group, the notorious ShinyHunters. 

While the size of the data breach is itself notable, the hack and subsequent leak of the European Commission’s data by two separate hacking groups highlights a growing trend of cybercriminals working together to extort their victims.

CERT-EU said that the breach originated on March 19 when hackers acquired a secret API key associated with the European Commission’s AWS account, following an earlier hack targeting the open-source security tool Trivy. The Commission inadvertently downloaded a copy of the compromised Trivy tool following the project’s recent breach, allowing the hackers to steal its secret API key and use that access to pivot to obtain data stored in the Commission’s AWS account.

While the service said it’s still analyzing the data published online, close to 52,000 files contain sent email messages. CERT-EU said the majority of these emails are automated with little to no content, but emails that bounced back with an error “may contain the original user-submitted content, posing a risk of personal data exposure.”

CERT-EU said it is already in contact with affected organizations. 

A spokesperson for the European Commission told TechCrunch that the body is closed until next week, and would respond to a request for comment then. 

A member of ShinyHunters did not respond to requests for comment. 

Besides the Trivy beach, TeamPCP has been linked to ransomware attacks and crypto-mining campaigns, says Aqua Security, which develops Trivy. The hackers have more recently been behind a systematic campaign of supply chain attacks compromising other open source security projects, according to Palo Alto Networks Unit 42.

By targeting developers with keys to access sensitive systems, the hackers “then have the ability to hold compromised organizations for ransom, demanding extortion payments,” Unit 42 wrote.