Tag: Exclusive

The final days of the Tesla Model X and S are here. All bets are on the Cybercab.

Posted on by faz_business


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.

Techcrunch event

San Francisco, CA
|
October 13-15, 2026

The Model X followed in fall 2015 and was famously described by Musk as the faberge 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 its 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 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 permit 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%.”



Meet the former Apple designer building a new AI interface at Hark

Posted on by faz_business


A secretive AI lab founded by serial entrepreneur Brett Adcock shared new details about what it believes is a novel marriage of model-building and hardware design that will change how humans interact with intelligent software.

The company said in a statement it would design multi-modal end-to-end models, their hardware, and their interfaces in tandem to deliver a “seamless end-to-end personal intelligence product.” The system will have a persistent memory of your life and can listen, see, and interact with the world in real time.

How that will be executed remains unclear outside the company, but Hark’s ambition is representative of Silicon Valley’s ongoing hunt for the killer app that will make AI a desired consumer product, not features kludged dubiously into existing digital platforms.

“My view is simple: today’s AI models aren’t nearly intelligent enough, they feel quite dumb, and the devices we use to access them are fundamentally pre-AI,” Adcock wrote in a January internal memo shared with TechCrunch. “We’re moving toward a world that looks more like sci-fi characters Jarvis or Her, with systems that anticipate, adapt, and genuinely care about the people using them.”

Details are intentionally sparse, but Hark points to Director of Design Abidur Chowdhury as a key hire. Previously an industrial designer at Apple credited with leading the design team behind the iPhone Air and other recent models, London-born Chowdhury left last fall after meeting with Adcock and buying into his vision for updating the way humans automate their lives.

In an exclusive interview with TechCrunch, Chowdhury declined repeated invitations to spill the beans on Hark’s roadmap, only saying that the public can anticipate a first release of the company’s AI models this summer. Asked about different approaches to working and living alongside AI, the designer did offer a few clues. 

 

Techcrunch event

San Francisco, CA
|
October 13-15, 2026

“What was very clear for me at the time is that the world is clearly changing, but we’re using the same devices…everything’s been designed around these existing platforms,” Chowdhury told. “Very few people are really going after what the future is. There’s so much that we could be doing if intelligence was at the base layer of everything we touched instead of becoming an app or a website at that upper layer.”

Chowdhury points to the awkwardness of everyday tasks of filling out forms, sharing information between devices, or the mundane tasks of booking travel or planning home renovation.

“Those are entire evenings of time where I have to plan…the anxiety of, you know, I spend my work day thinking about this in the back of my head, oh, I have to do this,” Chowdhury said. “We genuinely believe that all of the small tasks that pile up to be kind of gargantuan things today can be sort of automated from our lives.”

Chowdhury says the company knows what it is building, but can’t yet say how users will experience it. His comments suggest that wearables, like Meta’s Glasses, seem unlikely.

“I’m not the biggest believer in a lot of the wearable AI platforms that people are talking about right now,” Chowdhury said. “I don’t think it’s appropriate to put a layer between humanity and the interfaces we use in the world. I have similar discomfort with pins, or that kind of stuff that is going around with cameras.”

When generative AI first arrived on the scene, Chowdhury at first saw it as a flash in the pan, but successive generations of models convinced him that it would change his work. Hark, the word, means to pay attention, which Chowdhury says offers a thoughtful framing for the company’s mission.

“Traditional user experience always is about finding the simplest thing for everyone,” he told TechCrunch. “The future user experience will be finding the right thing for each individual. And I believe that can happen. But it requires a lot of work.”

The focus on elegance and simplicity for users echoes the high points of Apple’s product design, and naturally brings to mind Jony Ive, the legendary former Apple designer who is now developing AI native-hardware at OpenAI. A comparison that a Hark’s spokesperson declined to explore. 

Another parallel that comes to mind is how Elon Musk’s xAI work on advanced models dovetails with Tesla’s work on autonomous vehicles and humanoid robots.

There is similar corporate synergy between Adcock’s humanoid robotics company Figure and the new AI labs. Hark’s models are already being trained on Figure’s robots, although it is not clear to what end. A person familiar with the companies’ plans says there is no intention to combine them.

Hark employs 45 engineers and designers, including former Meta AI researchers and designers from Apple and Tesla, all of whom are working on the same campus that hosts Adcock’s other companies. Hark expects to begin using a new cluster of thousands of NVIDIA GPUs in April.

Now Hark, backed by $100 million in personal seed money from Adcock, will join the scramble for talent as the world’s biggest companies try to figure out the format that brings deep learning models into daily life — and at a time when frustration with the existing models for digital life is hitting a fever pitch.

“It just feels like there’s an opportunity for better, and I’ve not felt like that since the iPhone came up,” Chowdhury said.

The SEC drops its four-year-old investigation into EV startup Faraday Future

Posted on by faz_business


The Securities and Exchange Commission has closed its investigation into electric vehicle startup Faraday Future, despite SEC staff on the case recommending an enforcement action last year, TechCrunch has learned.

Four sources familiar with the investigation, who were granted anonymity to speak about the government case, told TechCrunch that the SEC informed the company and people involved in the probe about the closure this past week.

The dismissal of the case comes amid a historic drop in enforcement actions by the SEC, which only initiated four cases against publicly-traded companies in its 2025 fiscal year, a recent report shows. The SEC did not respond to an after-hours request for comment.

The investigation into Faraday Future lasted for nearly four years. The SEC was looking at whether the EV startup made “false and misleading statements” when it went public in a 2021 merger with a special purpose acquisition company (SPAC), and was also probing whether Faraday Future faked the sales of its first electric vehicles in 2023 — a claim that’s been made by at least three former employee whistleblowers.

The financial regulator sent the startup multiple subpoenas, regulatory filings from Faraday Future show. The SEC also took depositions of multiple former employees and executives in 2024 and 2025, three of the people familiar with the case have told TechCrunch.

In July 2025, Faraday Future revealed the SEC had sent the company and multiple executives — including founder Jia Yueting — letters known as “Wells Notices.” The SEC sends Wells Notices when staff working a case have decided to recommend the agency take enforcement action.

“We can now put all our energy into strategy execution. Over the past five years, we had to spend a great deal of time, effort, and money on cooperating with the investigation,” Jia said in a statement Sunday. Faraday Future said the SEC informed the company that it won’t take action against any of its executives, either.

Techcrunch event

San Francisco, CA
|
October 13-15, 2026

It’s not clear if Faraday Future ever responded to the Wells Notices sent last year. As recently as February, the company disclosed in regulatory filings that it had not. “The Company and executives plan to engage with the SEC to explain why enforcement action is not warranted,” Faraday Future wrote in such a filing last month.

The Department of Justice also sent Faraday Future requests for information after the SEC opened its investigation in 2022. Faraday Future has referred to this as an “investigation” in regulatory filings; the DOJ has never confirmed if it opened a full probe, and it did not respond to an after-hours request for comment.

It is rare for the SEC to not pursue an enforcement action after sending a Wells Notice. One study done at the Wharton School in 2020 showed that around 85% of targets who receive a Wells Notice wind up in court with the SEC.

The SEC investigated nearly every electric vehicle startup that went public in a SPAC merger over the last six years. In almost all of those cases, the agency reached a settlement with the startups. It dismissed an investigation into Lucid Motors in 2023, and as TechCrunch first reported in February, the SEC ended a probe into bankrupt EV startup Fisker late last year.

Origins of the investigation

Faraday Future was founded in California in 2014 by Jia, a businessman who at the time was running a booming tech conglomerate in China known as LeEco. It was one of many new companies trying to become the “next Tesla” or, optimistically, a “Tesla killer.”

Faraday snapped up talent from Tesla, other automakers, and also tech companies like Apple, and at one point employed as many as around 1,400 employees. But things got bumpy quickly. The company turned heads, in both good and bad ways, at the 2016 Consumer Electronics Show, with a flashy concept car and the lofty goal of being as disruptive as the iPhone.

The company revealed its first vehicle the following year: a luxury electric SUV called the FF91. By the end of 2017, though the company was nearly out of cash and had laid off or furloughed hundreds of workers. Jia’s company in China had collapsed, and he self-exiled to California as the government in his home country placed him on a debtor blacklist. (It was at this time that a close business associate to Jeffrey Epstein pitched the sex criminal on investing in Faraday Future, as well as other EV startups, as TechCrunch recently revealed. Epstein never invested.)

Faraday Future was rescued by an investment from major Chinese real estate conglomerate Evergrande. But that relationship fell apart quickly, too, with Evergrande walking away by the end of 2018 and Faraday Future laying off even more employees.

Jia nominally stepped aside as CEO in 2019 and also filed for personal bankruptcy to settle billions of dollars of LeEco debt he had personally guaranteed. But behind the scenes, he was still largely in charge of the company.

This became an issue when Faraday Future went public in 2021 and raised about $1 billion. Members of the newly-appointed public company board believed that Faraday’s executives had misrepresented Jia’s control over the day-to-day operations — especially after a short seller report was published that scrutinized Faraday Future — and formed a special committee to investigate.

That committee hired an outside law firm and a forensic accounting firm, and within the first few months it started reporting its findings directly to the SEC, the three people familiar with the investigation told TechCrunch.

Between January and April 2022, Jia was sidelined as a result of the board’s investigation, a senior VP named Matthias Aydt (who is now co-CEO with Jia) was placed on probation for six months, and another VP named Jerry Wang (who is Jia’s nephew) was suspended. (Wang ultimately resigned after “failure to cooperate with the investigation,” according to company filings, but is now back with Faraday Future.)

The committee’s work also showed that Faraday Future had, in the two years before it went public, survived in part on multi-million-dollar loans made to the company by low-level employees with connections to Jia — known as “related party transactions” in legal parlance.

On March 31, 2022, Faraday Future disclosed that the SEC had opened its investigation. The startup revealed the requests for information from the DOJ in June.

Dodging another bullet

Through the rest of 2022, and amid the early stages of the SEC investigation, employees and people close to Jia waged a campaign to regain control of the board and his company. This eventually resulted in death threats against some directors, who ultimately resigned, paving the way for people close to Jia to run the company once more.

Faraday Future finally delivered the first few FF91 SUVs in early 2023. Former employees have sued the company alleging that these were not true sales, and that the company had misled investors. The SEC investigators working the case subpoenaed Faraday Future about issues related to these sales, filings show.

Former executives and employees were initially deposed by the SEC in 2024, according to the people familiar with the investigation. The SEC sat some of them for longer depositions in the first half of 2025, the people said.

The Wells Notice sent in July 2025 said SEC staff had made “a preliminary determination to recommend that the Commission file an enforcement action against the Company alleging violations of various anti-fraud provisions of the federal securities laws.”

Specifically, the Wells Notice referenced “purported false or misleading statements” made during the SPAC merger process about “related party transactions” and Jia’s “role in the Company.” Jia, his nephew Wang, and two other unnamed employees also received Wells Notices.

Faraday Future is still trying to sell the FF91, but it has also recently changed its business in a few ways. The company is importing more affordable hybrid and electric vans from China. It also appears to be selling re-badged versions of Chinese robots, and turned a publicly-traded biotechnology company into a firm focused on crypto.

Those efforts have not stopped the company’s struggles. On Friday, the company announced it had received a warning from the Nasdaq that its stock price was under the minimum of $1, which could eventually lead to the company being de-listed.

This story has been updated with a statement from Faraday Future.

Mistral bets on ‘build-your-own AI’ as it takes on OpenAI, Anthropic in the enterprise

Posted on by faz_business


Most enterprise AI projects fail not because companies lack the technology, but because the models they’re using don’t understand their business. The models are often trained on the internet, rather than decades of internal documents, workflows, and institutional knowledge. 

That gap is where Mistral, the French AI startup, sees opportunity. On Tuesday, the company announced Mistral Forge, a platform that lets enterprises build custom models trained on their own data. Mistral announced the platform at Nvidia GTC, Nvidia’s annual technology conference, which this year is focused heavily on AI and agentic models for enterprise.

It’s a pointed move for Mistral, a company that has built its business on corporate clients while rivals OpenAI and Anthropic have soared ahead in terms of consumer adoption. CEO Arthur Mensch says Mistral’s laser focus on the enterprise is working: The company is on track to surpass $1 billion in annual recurring revenue this year.

A big part of doubling down on enterprise is giving companies more control over their data and their AI systems, Mistral says. 

“What Forge does is it lets enterprises and governments customize AI models for their specific needs,” Elisa Salamanca, Mistral’s head of product, told TechCrunch. 

Several companies in the enterprise AI space already claim to offer similar capabilities, but most focus on fine-tuning existing models or layering proprietary data on top through techniques like retrieval augmented generation (RAG). These approaches don’t fundamentally retrain models; instead, they adapt or query them at runtime using company data.

Mistral, by contrast, says it is enabling companies to train models from scratch. In theory, this could address some of the limitations of more common approaches — for example, better handling of non-English or highly domain-specific data, and greater control over model behavior. It could also allow companies to train agentic systems using reinforcement learning and reduce reliance on third-party model providers, avoiding risks like model changes or deprecation. 

Techcrunch event

San Francisco, CA
|
October 13-15, 2026

Forge customers can build their custom models using Mistral’s wide library of open-weight AI models, which includes small models such as the recently introduced Mistral Small 4. According to Mistral co-founder and chief technologist, Timothée Lacroix, Forge can help unlock more value out of its existing models. 

“The trade-offs that we make when we build smaller models is that they just cannot be as good on every topic as their larger counterparts, and so the ability to customize them lets us pick what we emphasize and what we drop,” Lacroix said. 

Mistral advises on which models and infrastructure to use, but both decisions stay with the customer, Lacroix said. And for teams that need more than guidance, Forge comes with Mistral’s team of forward-deployed engineers who embed directly with customers to surface the right data and adapt to their needs — a model borrowed from the likes of IBM and Palantir. 

“As a product, Forge already comes with all the tooling and infrastructure so you can generate synthetic data pipelines,” Salamanca said. “But understanding how to build the right evals and making sure that you have the right amount of data is something that enterprises usually don’t have the right expertise for, and that’s what the FDEs bring to the table.” 

Mistral has already made Forge available to partners, including Ericsson, the European Space Agency, Italian consulting company Reply, and Singapore’s DSO and HTX. Early adopters also include ASML, the Dutch chipmaker that led Mistral’s Series C round last September at a €11.7 billion valuation (approximately $13.8 billion at the time).

These partnerships are emblematic of what Mistral expects Forge’s main use cases to be. According to Mistral’s chief revenue officer Marjorie Janiewicz, these include governments who need to tailor models for their language and culture; financial players with high compliance requirements; manufacturers with customization needs; and tech companies that need to tune models to their code base.

Stripe, PayPal Ventures bet on India’s Xflow to fix cross-border B2B payments

Posted on by faz_business


Xflow, an Indian fintech startup, has secured backing from both Stripe and PayPal Ventures in a $16.6 million funding round. The investment comes as the company works to carve out a position in cross-border B2B payments, a market still dominated by banks and manual processes.

The Series A round was led by General Catalyst, with participation from existing investors Square Peg, Stripe, Lightspeed, and Moore Capital, while PayPal Ventures joined as a new backer. The all-equity round values the Bengaluru-based startup at $85 million post-investment and brings its total funding to more than $32 million to date.

Despite rapid digitization in domestic payments, cross-border B2B transfers for Indian exporters remain heavily reliant on banks, often with limited visibility into fees, settlement timelines, and the final amount received in rupees. The friction is particularly acute for larger exporters moving millions of dollars into India to fund salaries and local operations, creating an opening for fintech infrastructure players such as Xflow that promise greater transparency and speed in international money movement.

Founded in 2021, Xflow provides cross-border payment infrastructure for businesses ranging from exporters and SaaS firms to platforms and freelancers, enabling them to collect international payments, manage foreign exchange, and settle funds in India.

“Cross-border B2B payments were stuck in a different age compared to UPI,” co-founder Anand Balaji (pictured above, center) said in an interview, referring to India’s widely used instant domestic payments network, the Unified Payments Interface.

Balaji, who previously helped build out Stripe’s India business, founded Xflow with former Stripe colleagues Ashwin Bhatnagar (pictured above, right) and Abhijit Chandrasekaran (pictured above, left).

Last year, Xflow said it enabled Indian businesses to collect payments from more than 100 countries in over 25 currencies. It processed close to $1 billion in annualized cross-border payment volume last year, marking roughly 10-fold growth from the same period in 2024, Balaji told TechCrunch.

Techcrunch event

Boston, MA
|
June 9, 2026

According to the company, its customer base has expanded to about 15,000 businesses spanning SaaS firms, global capability centers (which are offshore units that multinationals operate in India), IT services exporters, freelancers, and fintech platforms.

Transaction sizes vary widely by segment, with global capability centers averaging about $1 million to $2 million per transaction, goods exporters around $30,000 to $40,000, and freelancers roughly $3,000, according to Balaji.

Xflow is positioning itself as a payments infrastructure provider rather than a direct payments application, offering APIs that allow platforms and exporters to embed cross-border money movement into their own products.

“We didn’t want to build the next Wise — we want to power the next thousand Wises,” Balaji said.

The startup has also introduced an AI-based foreign exchange tool to help finance teams optimize the timing of currency conversions. Xflow says the feature has generated incremental gains for some customers through data-driven foreign exchange decisions.

The tool allows businesses to set target conversion rates rather than accepting prevailing bank quotes. Balaji likened the feature to limit orders in trading — instructions to buy or sell only at a specified price.

“What we’ve added is the prediction layer and the ability to actually set a limit order,” he said. The model currently provides a three-day forecast with about 92% confidence, Balaji said, though TechCrunch could not independently verify that figure.

Xflow faces competition from banks that still dominate large cross-border B2B transfers, as well as fintech players such as Wise, Payoneer, and Skydo at the lower end of the market. But Balaji said the startup’s focus on high-value transactions and API-led infrastructure differentiates it from many rivals.

The startup plans to deploy the new capital toward building additional products on top of its core payments infrastructure and securing regulatory licenses in new markets, Balaji said. Xflow is preparing to roll out import capabilities in the coming months and is pursuing licenses in markets including Singapore, while already holding a payments license in Canada, even as it remains focused on India as its primary market.

Xflow said it has also received final authorization from the Reserve Bank of India for a Payment Aggregator–Cross Border (PA-CB) license covering both exports and imports. The startup has signed platform partnerships with Easebuzz and Drip Capital to embed its cross-border capabilities into their offerings.

Backing from Stripe and PayPal Ventures, Balaji said, has helped strengthen the startup’s credibility with banking and regulatory partners, even as it continues to work with multiple payment providers commercially.

The startup currently has about 65 employees as it scales its cross-border infrastructure business.

Qualcomm backs SpotDraft to scale on-device contract AI with valuation doubling toward $400M

Posted on by faz_business


As demand grows for privacy-first enterprise AI that can run without sending sensitive data to the cloud, SpotDraft has raised $8 million from Qualcomm Ventures in a strategic Series B extension to scale its on-device contract review tech for regulated legal workflows.

The extension values SpotDraft at around $380 million, the startup told TechCrunch, nearly double its $190 million post-money valuation following its $56 million Series B in February of last year.

Across regulated sectors, enterprises have moved quickly to test generative AI, but privacy, security, and data governance concerns continue to slow adoption for sensitive workflows — especially in legal, where contracts can include privileged information, intellectual property, pricing, and deal terms. Industry research has consistently flagged data security and privacy as key barriers to wider GenAI deployment in professional services, pushing vendors like SpotDraft to pursue architectures that keep core contract intelligence on the user’s device rather than routing it through the cloud.

At Qualcomm’s Snapdragon Summit 2025, SpotDraft demonstrated its VerifAI workflow running end-to-end on Snapdragon X Elite-powered laptops, executing contract review and edits offline while keeping the document on the local machine. SpotDraft said internet connectivity is still required for login, licensing, and collaboration features, but contract review, risk scoring, and redlining can run fully offline without sending documents to the cloud.

SpotDraft sees legal as an early proving ground for on-device enterprise AI, arguing that sensitive contracts often cannot be routed through external cloud models due to privacy, security, and compliance constraints.

“The future of how enterprise AI is going to be — right now, there’s got to be AI that is close to the document, which is privacy critical, latency sensitive, [and] legally sensitive, and those are the things that will move on device,” said Shashank Bijapur (pictured above, left), co-founder and CEO of SpotDraft, in an interview.

SpotDraft says VerifAI’s on-device capability extends beyond simply generating summaries, with the tool designed to apply playbooks and recommendations directly inside Microsoft Word, the way legal teams already work. “VerifAI will compare a contract against your guidelines, your playbooks, your prior policies,” said Madhav Bhagat (pictured above, right), co-founder and CTO of SpotDraft.

Techcrunch event

San Francisco
|
October 13-15, 2026

SpotDraft's VerifAI in Microsoft Word
SpotDraft’s VerifAI works in Microsoft WordImage Credits:SpotDraft

Bijapur told TechCrunch that the demand for on-device AI is emerging most clearly in tightly regulated sectors, including defense and pharma, where internal security reviews and data residency requirements can slow or block the use of cloud-based AI tools for sensitive documents.

On-device models have rapidly closed the gap with cloud-based systems, both in output quality and response times, Bhagat said. “Now we’ve come to a place where, in terms of eval, we are seeing as little as 5% difference between the frontier models, and some of these fine-tuned on device models,” he said, adding that speeds on newer chips are now “one-third of what we get in the cloud.”

Since its launch in 2017, SpotDraft said it has reached more than 700 customers, up from around 400 in February last year, and counts Apollo.io, Panasonic, Zeplin, and Whatfix among its users. The company said adoption is rising on its contract lifecycle management platform, with customers now processing over 1 million contracts annually, contract volumes growing 173% year-over-year, and nearly 50,000 monthly active users. It also expects 100% year-over-year revenue growth in 2026, after growing 169% in 2024 and posting a similar growth rate in 2025, though it did not share specific revenue figures.

SpotDraft plans to use the new capital to deepen its product and AI capabilities and expand its enterprise presence across the Americas, the EMEA region (Europe, Middle East, and Africa), and India, Bijapur said, adding that Qualcomm’s involvement extends beyond financing into joint development and go-to-market efforts for on-device deployments. The startup’s on-device workflow is currently available to a limited set of customers, and the founders expect it to expand more broadly as compatible AI PC hardware becomes more widely available.

“SpotDraft’s ability to deploy their proprietary models securely on-device using Snapdragon platforms represents a meaningful advancement for a privacy-critical industry,” said Quinn Li, senior vice president, Qualcomm Technologies, and global head of Qualcomm Ventures.

Bengaluru- and New York-based SpotDraft said it has a team of 300-plus employees, including 15–20 in the U.S., where COO Akshay Verma is based, and four to five in the UK, with the rest of the workforce in Bengaluru.

To date, the startup has raised $92 million, including the latest Qualcomm Ventures investment. Its earlier investors include Vertex Growth Singapore, Trident Growth Partners, Xeed VC, Arkam Ventures, and Prosus Ventures.

a16z pauses its famed TxO Fund for underserved founders, lays off staff

Posted on by faz_business


Andreessen Horowitz is pausing its Talent x Opportunity (TxO) fund and program, according to four sources familiar with the matter, including more than one founder in the program. 

The firm announced TxO in 2020 to support founders who do not have access to traditional venture networks. Many of TxO’s participants were women and minorities who, overall, receive very slim amounts of venture capital dollars.

The announcement of the fund came during the wave of support that underrepresented founders received in 2020 after the murder of George Floyd. The fund launched with $2.2 million in initial commitments, TechCrunch previously reported, with a16z co-founder Ben Horowitz and his wife, Felicia, matching up to an additional $5 million.

TxO provided founders with access to tech networks, a 16-week-long training program, and a $175,000 investment through a donor-advised fund managed by the nonprofit Tides Foundation. The program went on to support more than 60 companies (like the media brand Brown Girl Magazine, food tech Myles Comfort Foods, and the maternity tech Villie). 

TxO garnered some criticism when it launched because it’s technically structured as more of a nonprofit, rather than a traditional investment fund. Those investing in the fund are considered donors, and the money given is regarded as charity donations, rather than traditional limited partner investments.

Still, founders who participated in the program and spoke to TechCrunch said it provided them with invaluable support and opportunities to which they otherwise would not have access. Last year, TxO expanded to launch a grant program, providing $50,000 to three tech nonprofits that support underserved founders. 

TxO announced its — as of now — last cohort of the program in early March 2025. Founders who partook in the program received an email on October 16 from Kofi Ampadu, the partner at a16z who led TxO, announcing the program would pause. 

Techcrunch event

San Francisco
|
October 13-15, 2026

“When we launched TxO, the mission was clear: support talented, determined builders who are creating culture-shaping companies but may not have access to typical Silicon Valley networks and resources,” Ampadu’s email read, as seen by TechCrunch. “While that purpose has not changed, we are pausing our existing program to refine how we deliver on it.”

The rest of the email read:  

Over the past five years, we’ve experimented with different models for best serving founders — from virtual and in-person programming to curriculum design and funding structure. As we rethink what’s next, we’ll be applying everything we’ve learned and evolving how we support founders by integrating with a16z’s broader early-stage investing and company building strategy.  

TxO has backed more than 60 companies and nearly 100 founders. You have collectively raised tens of millions in follow-on capital and reached customers across culture and lifestyle. Founders from earlier cohorts now advise newer ones, and that peer support has strengthened the entire community.  

Thank you for being at the center of this community. Your progress is proof of what is possible. Stay tuned for what comes next. In the meantime, if you have any questions, please don’t hesitate to reach out directly.

Best regards,

Kofi

A16z confirmed to TechCrunch that the program was shutting down and that Ampadu alerted participants via email.

Members of the TxO staff team, which had at least three people, excluding Ampadu, were also let go, according to two sources, with the end of October being their last week. 

The fund’s application documents did not specifically call for founder diversity, except in terms of “cultural authenticity,” and also emphasized classic startup investment criteria like size of the market and ability to execute.  But the announcement of the fund back in 2020 made clear it was “for entrepreneurs who did not have access to the fast track in life but who have great potential. Their products can be non-tech or tech; they should be from underserved communities (all backgrounds welcome).”

Still, many in the startup world perceived TxO as an accelerator for diverse talent, and several people who spoke to TechCrunch pointed out that its hiatus comes as top names in tech eliminate, cut, reframe, or completely walk back on prior public commitments related to diversity, equity, and inclusion. The Trump administration has threatened legal and political ramifications for businesses supporting anything that could be seen as DEI. 

Others, however, noted that a16z is still interested in accelerator-type startup programs. Earlier this year, it launched Speedrun, a program that promises cohort grads up to $1 million of investment.

How Bill Gates’s fellowship program is adapting to global uncertainty

Posted on by faz_business


There’s plenty of uncertainty to go around this year, including a global trade war, shifting policy priorities, and an economy that’s starting to stumble. Breakthrough Energy, a climate tech organization founded by Bill Gates, has also been shifting in response.

The group always placed long bets, though it appears to be reappraising some of them. Its policy team was scrapped in March, for example, and it didn’t continue funding a publication that covered the climate tech world. Still, its investments in startups continue, as does its longest bet, a fellowship program for budding entrepreneurs.

Breakthrough Energy Fellows, as the program is called, is announcing a new cohort today, TechCrunch exclusively learned. It consists of 45 fellows at 22 different startups, and its makeup reveals how the program is evolving both in response to its own data and to global uncertainty.

“It’s the most global [cohort] that we’ve had to date. Fifty percent of the teams are based outside of the U.S.,” Ashley Grosh, vice president at Breakthrough Energy, told TechCrunch.

Grosh and her colleagues had to sift through around 1,500 applications and referrals, making the program more selective than the world’s top universities. Eleven teams are based in the U.S., six are in Asia, and the remainder are in Canada, Germany, the U.K., and South Africa.

Part of the international focus was driven by a new hub for the fellowship program in Singapore, which the organization opened in August 2024 with Temasek, the country’s investment fund, and Enterprise Singapore, a government agency. 

But it’s also a recognition that climate change, being a global problem, will require solutions from around the world. 

Techcrunch event

San Francisco
|
October 27-29, 2025

“What are local needs, right? What are the local challenges?” Grosh said. By way of example, she points to the fact that several cohort members are working on hydrogen.

In Asia, “there’s a lot of interest in the hydrogen economy,” Grosh noted. Circularity, or recycling materials back to their original form, or better, is also a priority for the region, given its role as a global factory and all the waste that entails.

The new cohort also has startups working on critical minerals, agriculture, and grid modernization.

Beyond its more global focus, the Breakthrough Energy Fellows program has also shifted its curriculum. Based on observations and feedback from previous cohorts, it is encouraging the new group to think early and often about the economics of the technology they’re developing. Using a framework called techno-economic analysis, they work with “business fellows” — often entrepreneurs with relevant experience — to determine whether and where their idea can find product-market fit. If not, they’ll be nudged to pivot.

“We were seeing a lot of companies come in thinking that they’re going to do one thing, and then they pivot,” Grosh said. “They’re more venture bankable once we’ve helped them through that pivot and validated it.”

Grosh said that nearly all of the teams from the previous four cohorts have raised follow-on funding, and one, Holocene, has already exited. “That’s a huge measure of success for us,” she said.

OpenAI reorganizes research team behind ChatGPT’s personality

Posted on by faz_business


OpenAI is reorganizing its Model Behavior team, a small but influential group of researchers who shape how the company’s AI models interact with people, TechCrunch has learned.

In an August memo to staff seen by TechCrunch, OpenAI’s chief research officer Mark Chen said the Model Behavior team — which consists of roughly 14 researchers — would be joining the Post Training team, a larger research group responsible for improving the company’s AI models after their initial pre-training.

As part of the changes, the Model Behavior team will now report to OpenAI’s Post Training lead Max Schwarzer. An OpenAI spokesperson confirmed these changes to TechCrunch.

The Model Behavior team’s founding leader, Joanne Jang, is also moving on to start a new project at the company. In an interview with TechCrunch, Jang says she’s building out a new research team called OAI Labs, which will be responsible for “inventing and prototyping new interfaces for how people collaborate with AI.”

The Model Behavior team has become one of OpenAI’s key research groups, responsible for shaping the personality of the company’s AI models and for reducing sycophancy — which occurs when AI models simply agree with and reinforce user beliefs, even unhealthy ones, rather than offering balanced responses. The team has also worked on navigating political bias in model responses and helped OpenAI define its stance on AI consciousness.

In the memo to staff, Chen said that now is the time to bring the work of OpenAI’s Model Behavior team closer to core model development. By doing so, the company is signaling that the “personality” of its AI is now considered a critical factor in how the technology evolves.

In recent months, OpenAI has faced increased scrutiny over the behavior of its AI models. Users strongly objected to personality changes made to GPT-5, which the company said exhibited lower rates of sycophancy but seemed colder to some users. This led OpenAI to restore access to some of its legacy models, such as GPT-4o, and to release an update to make the newer GPT-5 responses feel “warmer and friendlier” without increasing sycophancy.

Techcrunch event

San Francisco
|
October 27-29, 2025

OpenAI and all AI model developers have to walk a fine line to make their AI chatbots friendly to talk to but not sycophantic. In August, the parents of a 16-year-old boy sued OpenAI over ChatGPT’s alleged role in their son’s suicide. The boy, Adam Raine, confided some of his suicidal thoughts and plans to ChatGPT (specifically a version powered by GPT-4o), according to court documents, in the months leading up to his death. The lawsuit alleges that GPT-4o failed to push back on his suicidal ideations.

The Model Behavior team has worked on every OpenAI model since GPT-4, including GPT-4o, GPT-4.5, and GPT-5. Before starting the unit, Jang previously worked on projects such as Dall-E 2, OpenAI’s early image-generation tool.

Jang announced in a post on X last week that she’s leaving the team to “begin something new at OpenAI.” The former head of Model Behavior has been with OpenAI for nearly four years.

Jang told TechCrunch she will serve as the general manager of OAI Labs, which will report to Chen for now. However, it’s early days, and it’s not clear yet what those novel interfaces will be, she said.

“I’m really excited to explore patterns that move us beyond the chat paradigm, which is currently associated more with companionship, or even agents, where there’s an emphasis on autonomy,” said Jang. “I’ve been thinking of [AI systems] as instruments for thinking, making, playing, doing, learning, and connecting.”

When asked whether OAI Labs will collaborate on these novel interfaces with former Apple design chief Jony Ive — who’s now working with OpenAI on a family of AI hardware devices — Jang said she’s open to lots of ideas. However, she said she’ll likely start with research areas she’s more familiar with.

This story was updated to include a link to Jang’s post announcing her new position, which was released after this story published. We also clarify the models that OpenAI’s Model Behavior team worked on.



TDK backs Ultraviolette with $21M to take India-made electric motorcycles global

Posted on by faz_business


Two months ago, Indian electric motorcycle startup Ultraviolette expanded into 10 European countries. Now, fueled with $21 million in an all-equity round led by the corporate venture arm of Japanese electronics giant TDK Corporation, Ultraviolette is putting its expansion plans into overdrive.

The nine-year-old startup plans to grow its European footprint fourfold, enter other motorcycle-driven markets such as Latin America and Southeast Asia, and increase its portfolio to 14 models by early 2027. Ultraviolette’s global expansion follows the 2024 launch of its F77 Mach 2 flagship model and its second product, the F77 SuperStreet, in February.

Behind Ultraviolette are two childhood friends — CEO Narayan Subramaniam and CTO Niraj Rajmohan — who combined their expertise in mechanical engineering, automotive design, computer science, and electronics to electrify the mid-segment two-wheeler market.

The duo, which drew inspiration from Tesla, started Ultraviolette at a time when India’s electric two-wheeler market was dominated by low-speed models, mainly catering to commercial and utility needs. The early boom was driven by Chinese imports offering low-cost options, followed by a wave of homegrown startups and, more recently, legacy manufacturers entering the space.

Instead of becoming just another player in that race, the Ultraviolette co-founders set out to build electric motorcycles that could match the performance of 150cc to 800cc internal combustion engine sports bikes.

“We asked ourselves, if we have to make electric exciting in two-wheelers, what would it take? And that’s the objective with which we started,” said Rajmohan (pictured above, right) in an exclusive interview.

The Bengaluru-based startup took about four years from its inception in 2016 to unveil the first model in 2019. The startup went through multiple design iterations before finalizing the seventh version — hence the name F77. The commercial version debuted with a fixed battery pack to deliver over 186 miles of range and a top speed of 96 miles/hour with a 30kW peak power and up to 100 newton-meters of torque.

Techcrunch event

San Francisco
|
October 27-29, 2025

Ultraviolette has also unveiled the lightweight Shock Wave motorcycle as well as the Tesseract scooter, which features front and rear radars and cameras to enable an assisted-driving experience and blindspot detection. The scooter costs ₹145,000 ($1,650), while its motorcycles (ex-showroom) have a base price of ₹175,000 ($2,000) and goes up to $10,000.

Ultraviolette F77 Mach 2Image Credits:Ultraviolette

Ultraviolette’s vehicles come equipped with eSIM connectivity and feature predictive maintenance powered by a proprietary diagnostics system. Rajmohan said the system can detect even minor issues, such as when the chain needs lubrication. The startup offers an app that provides all these insights to consumers on the go.

The company has also established a manufacturing and assembly facility in Bengaluru’s Electronics City, with a capacity of 30,000 units. Today, the company handles everything in-house from developing embedded software and battery management systems to motor controllers and even battery manufacturing. About 500 people work at Ultraviolette, including 200 in corporate functions and R&D.

Ultraviolette’s business model was shaped in part by Tesla owners. The co-founders spent time talking to Tesla owners in the U.S., who were among the first ones to buy the Model S in 2015, to learn what made the that car different from other EVs of its time.

“These Tesla cars were very special, as owning them was seen as progressive. It was more of a lifestyle statement,” Rajmohan told TechCrunch.

The co-founders brought that sentiment to Ultraviolette’s design and branding, aiming to make it a global company from day one. As Rajmohan explained, the word “violet” is pronounced similarly in over 30 European languages, while “ultra” signals something cutting-edge. Reinforcing that ambition, the startup pursued European certification for all its vehicles even before entering the market.

This is unlike other Indian electric two-wheeler manufacturers, which have tried to cater to local demand. India accounts for nearly 40% of global motorcycle sales — although most of those are powered by internal combustion engines.

Expanding beyond India makes strategic sense for Ultraviolette, given the domestic EV market remains relatively underpenetrated — with adoption at just 7.66%, compared to the global average of 16.48%, according to a recent report by government-backed think tank NITI Aayog. While India aims to reach 30% EV penetration by 2030, progress so far suggests that it may be an ambitious target.

EV penetration Rate — Global and IndiaImage Credits:NITI Aayog

India is also a price-sensitive market, where two-wheelers are typically not discretionary purchases, but essential and affordable modes of daily transportation. As a result, selling high-end variants at scale in the country could be a challenge for Ultraviolette — at least initially.

“We were very clear that what we’re doing is, we’re working toward segments which are more universal in nature,” Rajmohan said.

What’s next?

Ultraviolette’s manufacturing plant in BengaluruImage Credits:Ultraviolette

Ultraviolette plans to expand the capacity of its Bengaluru production facility to up to 60,000 units and add a larger location to scale to about 300,000 units by early next year. Ultraviolette operates 20 stores across 20 Indian cities and plans to grow to around 100 by March next year. About 50 of those stores — one per city — are expected to open by the festive season later this year.

Rajmohan told TechCrunch the startup is working on expanding its European presence, where it has 40 dealers.

“Next year is where the scale-up happens in Europe,” he said.

The startup also plans to start its pilot in Latin America and Southeast Asia next year and go to markets including the U.S. and Japan later.

Ultraviolette has sold more than 3,000 motorcycles in India and has projected to sell up to 10,000 later this year. It has also targeted over $50 million in revenue by the end of this financial year.

The new funding saw participation from Ultraviolette’s existing investors Zoho Corporation and Lingotto (previously Exor Capital). To date, it has raised around $75 million in funding and counts Qualcomm Ventures, Exor, and TVS Motor among its other key investors.