Nvidia CEO Jensen Huang says market got it wrong about DeepSeek’s impact
Nvidia founder and CEO Jensen Huang said the market got it wrong when it comes to DeepSeek’s technological advancements and its potential to negatively impact the chipmaker’s business.
Instead, Huang called DeepSeek’s R1 open source reasoning model “incredibly exciting” while speaking with Alex Bouzari, CEO of DataDirect Networks, in a pre-recorded interview that was released on Thursday.
“I think the market responded to R1, as in, ‘Oh my gosh. AI is finished,’” Huang told Bouzari. “You know, it dropped out of the sky. We don’t need to do any computing anymore. It’s exactly the opposite. It’s [the] complete opposite.”
Huang said that the release of R1 is inherently good for the AI market and will accelerate the adoption of AI as opposed to this release meaning that the market no longer had a use for compute resources — like the ones Nvidia produces.
“It’s making everybody take notice that, okay, there are opportunities to have the models be far more efficient than what we thought was possible,” Huang said. “And so it’s expanding, and it’s accelerating the adoption of AI.”
He also pointed out that, despite the advancements DeepSeek made in pre-training AI models, post-training will remain important and resource-intensive.
“Reasoning is a fairly compute-intensive part of it,” Huang added.
Nvidia declined to provide further commentary.
Huang’s comments come almost a month after DeepSeek released the open source version of its R1 model which rocked the AI market in general and seemed to disproportionately affect Nvidia. The company’s stock price plummeted 16.9% in one market day upon the release of DeepSeek’s news.
Nvidia’s stock closed at $142.62 a share on January 24, according to data from Yahoo Finance. The following Monday, January 27, the stock dropped rapidly and closed at $118.52 a share. This event wiped $600 billion off of Nvidia’s market cap in just three days.
The chip company’s stock has almost fully recovered since then. On Friday the stock opened at $140 a share, which means the company has been able to almost fully regain that lost value in about a month. Nvidia reports its Q4 earnings on February 26 which will likely address the market reaction more.
Meanwhile, DeepSeek announced on Thursday that it plans to open source five code repositories as part of an “open source week” event next week.
Keep reading the article on Tech Crunch
Report: OpenAI plans to shift compute needs from Microsoft to SoftBank
OpenAI is forecasting a major shift in the next five years around who it gets most of its computing power from, The Information reported on Friday.
By 2030, OpenAI expects to get three-quarters of its data center capacity from Stargate, a project that’s expected to be heavily financed by SoftBank, one of OpenAI’s newest financial backers. That represents a major shift away from Microsoft, OpenAI’s biggest shareholder, who fulfills most of the startup’s power needs today.
The change won’t happen overnight. OpenAI still plans to increase its spending on Microsoft-owned data centers in the next few years.
During that time, OpenAI’s overall costs are set to grow dramatically. The Information reports that OpenAI projects to burn $20 billion in cash during 2027, far more than the $5 billion it reportedly burned through in 2024. By 2030, OpenAI reportedly forecasts that its costs around running AI models, also known as inference, will outpace what the startup spends on training AI models.
Keep reading the article on Tech Crunch
Norway’s 1X is building a humanoid robot for the home
Norwegian robotics firm 1X unveiled its latest home robot, Neo Gamma, on Friday. The humanoid system will succeed Neo Beta, which debuted in August. Like its predecessors, the Neo Gamma is a prototype designed for testing in the home environment. Images of the robot show it performing a number of household tasks like making coffee, doing the laundry, and vacuuming.
1X says the bipedal robot is set to step outside the lab, with limited in-home testing, though the company is quick to add that the Gamma is a long way from commercial scaling and deployment.
Neo Gamma represents a softer side of the humanoid industry — both figuratively and literally. 1X has built the robot to be welcoming, with a friendlier design and a suit made of knitted nylon. The latter is designed to reduce potential injuries that might arise from robot-to-human contact.
Neo Gamma arrives amid a sea of humanoids from companies like Agility, Apptronik, Boston Dynamics, Figure, and Tesla. While firms like Figure already have their robotic systems operating in a mock home environment within their lab, all have prioritized warehouse and factory deployment. 1X’s home-first approach makes it unique among its direct peers.
Home robots have always been a tricky proposition. Beyond robotic vacuums produced by companies like iRobot, none have meaningfully penetrated the market. This isn’t from lack of trying — the technology simply isn’t there.
Home robots need to be useful, reliable, affordable, and significantly safer than their industrial counterparts. This is doubly the case given that age-tech is likely to be one of home humanoids’ key targets. As the average age of the population rises, independent living for older adults will become an increasingly important technology target.
Along with a softer shell, 1X points to advances in the Gamma’s on-board AI system as a key element in designing a safer robot. These systems need to be extremely aware of their surroundings so as to avoid causing potential harm to people or property. Teleoperation is an important part of the safety conversation, as well. While full autonomy is the end goal for most, it’s important that humans be able to take control of the system in a pinch, especially in the home.
Beyond its unique focus, 1X first crossed the radar of many in the industry when OpenAI was announced as an early backer. For many, the notion of embodied intelligence — AI with a physical presence — is the next logical step for the white-hot world of generative AI. OpenAI has since hedged its bets in the humanoid space, with both an investment in a competitor, Figure, as well as numerous rumors surrounding the ChatGPT maker’s own in-house robotics ambitions.
Generative AI has an important role to play with humanoids, including the creation of more natural person-to-robot language interactions. Much like Figure, 1X has been building its own in-house models designed to improve both the robot’s speech and body language. It’s unclear how many of Gamma’s new and improved features are a result of the company’s work with OpenAI or its January acquisition of Bay Area startup, Kind Humanoid.
1X has not disclosed how many Neo Gammas have been — or will be — produced over the course of the beta robot’s life. The product videos accompanying Friday’s launch, meanwhile, are best viewed as proof of concept of how one of Neo’s creations might behave in a home setting.
While we’re seeing the first humanoid deployments move beyond the pilot stage in industrial settings, these systems have a long way to go in terms of pricing, reliability, safety, and functionality before we can have a serious conversation about bringing them home.
Keep reading the article on Tech Crunch
Sakana walks back claims that its AI can dramatically speed up model training
This week, Sakana AI, an Nvidia-backed startup that’s raised hundreds of millions of dollars from VC firms, made a remarkable claim. The company said it had created an AI system, the AI CUDA Engineer, that could effectively speed up the training of certain AI models by a factor of up to 100x.
The only problem is, the system didn’t work.
Users on X quickly discovered that Sakana’s system actually resulted in worse-than-average model training performance. According to one user, Sakana’s AI resulted in a 3x slowdown — not a speedup.
What went wrong? A bug in the code, according to a post by Lucas Beyer, a member of the technical staff at OpenAI.
“Their orig code is wrong in [a] subtle way,” Beyer wrote on X. “The fact they run benchmarking TWICE with wildly different results should make them stop and think.”
In a postmortem published Friday, Sakana admitted that the system has found a way to — as Sakana described it — “cheat” and blamed the system’s tendency to “reward hack” — i.e. identify flaws to achieve high metrics without accomplishing the desired goal (speeding up model training). Similar phenomena has been observed in AI that’s trained to play games of chess.
According to Sakana, the system found exploits in the evaluation code that the company was using that allowed it to bypass validations for accuracy, among other checks. Sakana says it has addressed the issue, and that it intends to revise its claims in updated materials.
“We have since made the evaluation and runtime profiling harness more robust to eliminate many of such [sic] loopholes,” the company wrote in the X post. “We are in the process of revising our paper, and our results, to reflect and discuss the effects […] We deeply apologize for our oversight to our readers. We will provide a revision of this work soon, and discuss our learnings.”
Props to Sakana for owning up to the mistake. But the episode is a good reminder that if a claim sounds too good to be true, especially in AI, it probably is.
Keep reading the article on Tech Crunch
Despite recent layoffs, Meta is expanding in India
Meta made headlines last month for announcing plans to cut 5% of its employees, controversially deeming them “low performers.” But the job cuts aren’t holding Meta back from expanding in certain geographic areas.
Meta is setting up a new site in the country’s tech hub of Bengaluru (formerly known as Bangalore), multiple Meta employees posted on LinkedIn this month.
Meta is currently hiring for 41 positions there, according to its careers webpage, most of which were posted over the last month. The positions are split between software or machine learning engineer jobs, and roles focused on designing chips for Meta’s data centers.
Meta Bengaluru is looking for an “experienced Engineering Director to build and lead our engineering team in India,” one of Meta’s job ads, which was posted three weeks ago on LinkedIn, reads.
The engineering director in Bengaluru will be responsible for designing a strategy to hire and build founding engineering teams, plus help create “a vision for engineer teams in India,” the job ad states.
The new center is part of Meta’s Enterprise Engineering team, according to a Meta employee’s LinkedIn post. That team focuses on custom internal Meta tools, rather than on Meta’s best-known products like Facebook and Instagram.
While Meta has several existing offices in India, including Bengaluru, Hyderabad, Gurgaon, New Delhi, and Mumbai, most have fewer job openings and they’re mainly for non-engineering roles. Only 1 of the 12 available positions at the other locations is engineering-related, Meta’s careers page shows.
A Meta spokesperson in India told TechCrunch that it is recruiting for a “small number of engineering positions in Bengaluru.”
“We regularly update our location strategies to support our long-term investments,” the spokesperson said.
The 41 positions in Bangalore are a small proportion of Meta’s global job postings, which currently total over 1,700. But it represents a shift for Meta, which has not traditionally used India as an engineering hub — those positions have historically been based in North America and Europe.
In one example from 2022, an Indian software engineer made headlines after he was laid off by Meta just two days after relocating to Canada for the job.
Meta CEO Mark Zuckerberg has said that Meta intends to backfill the jobs it cut during its most recent round of layoffs.
Keep reading the article on Tech Crunch
Fintech founder Charlie Javice’s criminal trial has begun
The criminal trial against fintech startup founder Charlie Javice began on Friday, with lawyers laying out their opening arguments, Reuters reported.
Lawyers reiterated their original claims and defenses from the lawsuit filed by JPMorgan Chase against Javice in December of 2022. The financial services giant alleges that Javice had helped “fake millions of customers in order to induce the bank to buy her company,” student financial planning aid startup Frank, for $175 million. That charge was also the root of an SEC complaint, which charged that Javice “made numerous misrepresentations” about Frank’s purported millions of users to entice JPMorgan.
JPMorgan claims that it found out about the alleged fraud when more than 70% of marketing test emails to a list of Frank’s customers bounced back.
Javice’s attorneys claim that JPMorgan did ample diligence and this suit is a result of buyer’s remorse due to a government change in the way financial aid forms are filled out with alleging fraud a way to get out of the deal.
Javice, now 32 years old, could be sentenced to years in prison if she’s convicted of deception and creating fake data.
Want more fintech news in your inbox? Sign up for TechCrunch Fintech here.
Want to reach out with a tip? Email me at [email protected] or send me a message on Signal at 408.204.3036. You can also send a note to the whole TechCrunch crew at [email protected]. For more secure communications, click here to contact us, which includes SecureDrop and links to encrypted messaging apps.
Keep reading the article on Tech Crunch
Apple takes on recipe apps with Apple News+ Food
Recipe app developers just got new competition. On Friday, Apple introduced a soon-to-launch feature for Apple News+ subscribers called Apple News+ Food, a new section that will allow users to search, discover, save, and easily cook recipes from dozens of existing News+ publishing partners.
It’s set to roll out as part of iOS 18.4 and iPadOS 18.4 in April, but only in the U.S., U.K., Canada, and Australia.
Instead of building a stand-alone recipe app that could import content from all over the web — like recipes from blogs or TikTok videos — Apple News+ Food will only focus on recipes offered by Apple News+ publishers.
At launch, Apple aims to have north of 30 publishers on board, up from the 20 it’s currently testing. Existing partners include well-known brands like Allrecipes, Bon Appétit, Food & Wine, Good Food, Serious Eats, Epicurious, Good Housekeeping, Better Homes & Gardens, Southern Living, Delish, Real Simple, Country Living, and others. Tens of thousands of recipes will be available through the Apple News+ Food service, the company notes.
The new experience lets Apple’s publishing partners get their content in front of more consumers at a time when Google’s ability to refer direct traffic to their websites continues to decline.
iPhone and iPad users will be able to find a new Food section by scrolling down in the Apple News app’s Today feed. Here, they’ll find a featured recipe, curated by Apple’s editorial team, followed by a collection of food and dining-related stories, a broader recipe collection, plus links to the Food+ recipe catalog and their own saved recipes.
Apple notes that select food stories and recipes will also be available for users who do not subscribe to Apple News+.
The Apple News+ Food subscription service can be accessed in multiple ways.
You can either tap on the “More food” link from the Food section in the app’s Today tab or you can tap on the link to “Food” from the Following tab. (The latter is a more direct method if you want to bypass reading the news articles and go straight to the recipes.)
In the Food+ section, users will see the featured recipe, which is updated daily, alongside an expanded set of recommended stories related to their interests. That personalization improves the more users engage with the app.
Other curated sections include those that link to your saved recipes or other types of recipe collections, like those from certain publishers, a selection of popular recipes, or those focused on some type of theme — like healthy eating or weeknight chicken dinners, for example.
As users browse the recipes, they can choose to save a recipe directly to the News app for later reference.
If users are searching for something specific, they can look through Apple’s News+ Food’s recipe catalog, tapping on buttons to narrow searches by various filters like “dinner,” “easy,” “vegetarian,” “under 30 minutes,” and more. Filters are also available for searching across your saved recipes.
The recipes themselves are formatted to be clutter and ad-free, as well as easy to read — an experience that’s far less common on today’s web.
Key information — including the ingredients, steps, description, cooking time, servings, and more — is pulled out and featured in a clear format that highlights a photo of the dish and links back to the publisher’s website.
Other features Apple added also come in handy. One lets you tap on an ingredient to see the amount needed without having to scroll back to the ingredients list. Another lets you tap on the cooking time in the recipe’s instructions to automatically start a timer on your iPhone or iPad.
A dedicated cooking mode is available, too, which displays the recipe in full screen with larger text so you can follow instructions with minimal tapping and scrolling. In this mode, the screen will stay on, even if your device is normally set to turn off the screen after a period of time.
One thing Apple News+ Food is missing, however, is the ability to add your own recipes or those saved from elsewhere on the web, as well as any tools to import or export recipes to and from other apps. You also can’t save recipes directly from social media, though many home chefs today find recipes on places like TikTok and Instagram Reels.
With the launch of Apple News+ Food, the tech giant continues to inch its way into the mobile app ecosystem where it competes with third-party developers who help the company generate revenue through App Store purchases. Recent additions to the Apple app lineup over the past year or so include the party-planning app Invites, iOS 18’s new Passwords app, the Sports app, and the mobile Journal, for example.
Unlike independent developers, Apple can afford to launch new apps that don’t have to be supported by a business model other than continued iPhone sales. This puts smaller and indie developers at a distinct disadvantage.
In the case of Apple News+ Food, publishers weren’t additionally compensated for their recipes, TechCrunch understands. Instead, the experience is an extension of Apple’s existing relationship with its partners, where the iPhone maker generates revenue by selling ads within the publishers’ articles for a 30% cut of sales.
The new service requires an Apple News+ subscription, which is $12.99 per month in the U.S., £12.99 in the U.K., $16.99 in Canada, and $19.99 in Australia. That includes access to over 400 magazines, newspapers, and digital publishers.
Keep reading the article on Tech Crunch
The Vision Pro is getting Apple Intelligence in April
Apple Intelligence is heading to the Vision Pro as part of an upcoming operating system update. Apple confirmed on Friday that its generative AI platform will arrive on the extended reality headset as part of visionOS 2.4. A beta version of the software is currently available for developers. The public version is set for an April release.
Like the iPhone and Mac before it, the Vision Pro will receive Apple Intelligence updates in waves. The first set includes several familiar offerings, focused primarily on generating text and images. The company sees the addition of features like Rewrite, Proofread, and Summarize as key components for on-device workflow.
It’s worth keeping in mind that Apple has framed Vision Pro as a “spatial computing” device since the outset. For all the video, gaming, and other entertainment features, the company has sought to set the system apart from its extended reality predecessors by positioning it as an extension of desktop computing — or, as TechCrunch framed it in our review, “The infinite desktop.”
As it stands, composing text is a mixed bag on the headset. The default typing method requires the wearer to look at a letter, before pinching two fingers together to select. While well implemented, it’s cumbersome when faced with writing more than a word or two at a time. Voice addresses this bottleneck to a degree, and Apple’s recent AI-powered Siri supercharge bodes well for the smart assistant’s Vision Pro future.
Apple is banking on the combination of voice dictation and generative AI writing tools to deliver a smoother experience to convince more Vision Pro users to incorporate the headset into more of their existing workflows. At the very least, features like Message Summaries and email Smart Reply streamline interaction with different apps, without taking the user away from a given task.
Image Playground is the other big piece of the puzzle, bringing image generation to the wearable display as part of the visionOS 2.4 update. The feature is integrated directly into the visionOS Photos app, allowing users to create images through verbal prompts.
All of the above features have previously been rolled out on iOS, macOS, and iPadOS. There are no new Apple Intelligence features specific to Vision Pro arriving in this update.
Along with visionOS 2.4, Apple has also launched a Vision Pro iPhone app arriving with iOS 18.4, which is also now in beta. The app serves a few different purposes. Foremost is the ability to browse visionOS content, like TV shows and movies, which can then be transferred onto the headset. This feature appears to be, in part, a response to the limitations of wearing the headset, both in terms of personal comfort and battery life. If you’re going to be scrolling through content, you might as well do it from the comfort of your iPhone.
When an iPhone is unlocked and within proximity of the headset, the new app can also be used to manage guest accounts. The Vision Pro will prompt its owner when someone is attempting to sign in as a guest. A streaming image of the guest’s in-headset view is accessible through the new app, as well.
Keep reading the article on Tech Crunch
Augury and Hightouch joined the unicorn club
Welcome to Startups Weekly — your weekly recap of everything you can’t miss from the world of startups. Want it in your inbox every Friday? Sign up here.
Startup life is a story of births and deaths. This week confirmed this, and it confirmed that unicorn rounds are not dead.
Most interesting startup stories from the week
As happens often in the startup world, it’s been a week of beginnings and endings.
Thinking machines: Thinking Machines Lab, the new AI startup of former OpenAI CTO Mira Murati, came out of stealth this week, without confirming how much it may have raised from VCs. Meanwhile, Ilya Sutskever’s startup Safe Superintelligence (SSI) is reportedly raising around $1 billion.
Humane killing: Humane, the hardware startup behind the flawed AI Pin, announced that most of its assets have been acquired by HP for $116 million.
Money back guarantee: San Francisco-based startup Future Family launched a new IVF insurance product in the United States that could refund prospective parents if treatment doesn’t succeed after two rounds.
Low battery: Former Silicon Valley darling Nikola Corp. filed for bankruptcy this week. The hydrogen electric trucking startup failed to find a buyer or secure additional funds to maintain operations.
Most interesting VC and funding news this week
We thought the days of counting unicorns were behind us, but not so much. Plus, legal AI is hot, and looming budget cuts inspired a new biotech venture fund.
Quadricorn: Austin-based defense startup Saronic raised a $600 million Series C to build an autonomous ship factory, also quadrupling its valuation to $4 billion from its last round.
New horn: Six months after raising a $150 million Series C at a $1.25 billion post-money valuation, AI coding startup Codeium is in talks to raise a new round led by Kleiner Perkins at a $2.85 billion valuation, sources told TechCrunch.
Joining the club: Hightouch, a startup co-founded by a former engineering manager at Segment that provides AI-powered marketing tools, closed an $80 million Series C round at a $1.2 billion valuation.
Augury, whose AI-based hardware detects malfunctions in factory machines, secured $75 million in equity funding as part of a Series F round that it is still closing. CEO and founder Saar Yoskovitz said this was an up-round valuing the startup at over $1 billion.
After the billion: After raising $1 billion to date since 2015, fintech Varo closed on $29 million out of the $55 million Series G it has been hoping to raise. Its CEO and founder, Colin Walsh, also recently announced he was stepping down.
Halfway: Sanas, which leverages AI to change call center workers’ accents in real time, closed a $65 million funding round, valuing the company at over $500 million.
Legal AI: AI-powered legal tech startup Luminance raised $75 million in a Series C funding round that follows several other deals closed by competitors, confirming the space is heating up.
New fund: Amid uncertainty on funding from the National Institutes of Health (NIH), Altitude Lab’s Pre-seed Venture Fund will consider investing $100,000 to $250,000 in biotech startups that were qualified for Small Business Innovation Research (SBIR) grants from the NIH.
Last but not least
Andreessen Horowitz may have $45 billion in assets under management, but Marc Andreessen isn’t “chomping at the bit to take the firm public.” Speaking on this week’s Invest Like the Best podcast, a16z’s co-founder discussed his goal of building it into an enduring company.
Keep reading the article on Tech Crunch
Crypto exchange Bybit says it was hacked and lost around $1.4B
Crypto exchange Bybit announced on Friday that “a sophisticated attack” led to the theft of Ethereum (ETH) from one of the company’s offline wallets.
Bybit’s chief executive and co-founder Ben Zhou said in a livestream that the hackers stole around 401,346 ETH, which at the time of the theft amounts to about $1.4 billion.
Both crypto security firm Elliptic, as well as crypto security researcher ZachXBT, the total amount of ETH stolen is worth around $1.4 billion, making this the largest known theft of crypto in history. The previous highest crypto breaches were the hacks against the Ronin Network and Poly Network, which resulted in the loss of $624 million and $611 million, respectively, according to data collected by Rekt, a site that tracks web3 and crypto breaches.
“In fact, it may even be the largest single theft of all time,” Tom Robinson, Elliptic’s co-founder and chief scientist told TechCrunch, referring to any kind of theft, not just data breaches.
Prior to Bybit’s breach, the withdrawal of around $1 billion from the Central Bank of Iraq is said to be the largest bank robbery of all time, according to the financial news site World Finance.
Zhou wrote on X that the hacker “took control” of one of the company’s cold wallets, referring to a digital wallet that stores cryptocurrency but in theory isn’t connected to the internet, and transferred funds to a “warm” wallet, which is online.
When reached for comment, Bybit spokesperson Tony Au referred to Zhou’s public posts. In one post, Zhou wrote that the company is “solvent” and “can cover the loss” even if it can’t recover the stolen funds.
Bybit, which is based in Dubai, United Arab Emirates, had an estimated total assets of $16 billion as of last week, according to CoinMarketCap.
To put things in perspective, in all of 2024, the total amount of crypto stolen by hackers was around $2.2 billion, according to blockchain tracking firm Chainalysis. And, in 2023, it was around $2 billion, according to multiple estimates.
Keep reading the article on Tech Crunch
The real reason why oil and gas companies are bullish on carbon capture
Two years ago, oil and gas company Occidental bought carbon capture startup Carbon Engineering. The transaction was hailed as a win-win: A climate tech company scored a significant exit, and a fossil fuel company gained a foothold in a sector that could be worth up to $150 billion by 2050.
Now we have a better idea why Occidental was keen to pick up the pricey technology: They want to use it to pump more oil.
Previously, the company had said it would use the technology to zero out its climate impact. Yet on Occidental’s earnings call this week, CEO Vicki Hollub changed the tune, saying that injecting CO2 into wells to force out more oil was imperative to boosting oil production.
“Taking CO2 out of the atmosphere is a technology that needs to work for the United States, and President Trump knows the business case for this,” Hollub said. The Verge was the first to report on the comments.
Hollub compared using CO2 in enhanced oil recovery to fracking, the technology that sent U.S. oil and gas production skyrocketing.
But direct air capture, the technique used by Carbon Engineering to draw CO2 out of the atmosphere, remains expensive at $600 to $1,000 per metric ton. The Inflation Reduction Act, though, provides some significant incentives for using captured CO2 in enhanced oil recovery, up to $130 per metric ton in 2026 if the gas remains permanently stored underground. That’s not enough to make the practice attractive on its own, but coupled with carbon credit sales, Occidental expects it can turn a profit by the end of the decade.
The Trump administration has been working to dismantle climate-related government incentives, especially the Inflation Reduction Act. But with support from companies like Occidental and ExxonMobil it’s possible that the tax credits could survive.
Carbon capture has a long and tangled history with fossil fuel companies. They first started pumping oil into dwindling wells in the 1970s, though the CO2 came from underground deposits. In the early 1980s, pipelines started stretching out from Texas, but low oil prices prevented the technique from being widely used.
About a decade ago, NRG Energy took advantage of rising oil prices to build the country’s first carbon capture facility attached to a coal-fired power plant. Called Petra Nova, the small installation was designed to capture about a third of one boiler’s carbon dioxide and use that CO2 to boost production at a flagging oilfield southwest of Houston.
It worked, though not as well as expected. Production rose from around 300 barrels per day to 6,000 barrels, a significant bump but half of what had been forecasted. NRG shut down Petra Nova in 2020 as oil prices cratered early in the pandemic and sold it to JX Nippon three years later.
Oil prices have since recovered, but enhanced oil recovery using CO2 remains unattractive in part because there isn’t enough of the gas readily available — at least, not enough to raise production by the 50 billion to 70 billion barrels that Hollub predicts the technology will unlock.
Direct air capture could easily provide enough CO2. Humans have been pumping gigatons’ worth of the gas into the air by burning fossil fuels over the last century and a half. It’s possible that carbon captured from the air could be used to make oil carbon negative, meaning the process of drilling the oil stores more carbon than burning it releases, though the concept needs to be studied further.
It’s hard to know whether federal incentives for direct air capture will survive the next four years. But of all the tax credits in the Inflation Reduction Act, they might have the best chance thanks to oil companies’ desire to continue business as usual.
Keep reading the article on Tech Crunch
How automotive exec Crystal Brown founded CircNova, an AI drug discovery biotech
Tiny Michigan biotech startup CircNova has raised a $3.3 million seed round for its technology that uses AI to target “circular RNA.” The development holds promise as a new method to quickly develop therapies for conditions that currently have no drug treatments.
The new funding is also a victory lap for co-founder and CEO Crystal Brown, who took an unconventional path to becoming a biotech founder.
RNA, or ribonucleic acid, is a key molecule that helps convert genetic information into proteins. Circular RNA is a relatively newly discovered class of such structures that form a circle rather than a strand. It regulates critical biological processes, and the hope is that therapies based on these molecules will be able to target complex health issues.
CircNova has developed a “proprietary AI engine that allows us to identify, design, and then produce novel, non-coding, circular RNAs,” Brown told TechCrunch.
It’s an AI engine similar to Google’s DeepMind AlphaFold, in that it also uses deep learning AI — not some kind of large language model — to generate and analyze new circular RNA for therapeutic use.
CircNova has not only its NovaEngine, which it says is the first in the world to be able to predict circular RNA structures, but it also has a wet lab. That means its AI engine produces the actual physical molecules themselves, which can then be validated and researched in collaboration with the University of Michigan, Brown said.
“We can reverse engineer. We can go from sequence to structure. We can go from structure to sequence when developing the molecule,” she says.
The goal is to “treat diseases we haven’t treated so far, things like ovarian cancer, triple-negative breast cancer, neurodegenerative diseases, rare genetic diseases,” she describes.
The tech is based on the work of CircNova co-founder Joe Deangelo, the startup’s chief scientific officer and previous CEO of biotech Neochromosome as well as the former CSO of Apex Bioscience. Investor William Grenawitzke is chief business officer and the startup’s third co-founder.
Lessons from a failed startup
Brown seems like an unlikely founder of such a company because until about seven years ago, her career had been in the automotive manufacturing industry.
She thought she was climbing the ladder to become a “C-suite automotive executive” when a friend of hers introduced her to a CEO running a life science startup. The startup CEO was looking for a business manager.
Curious, Brown offered to keep the books part-time, which evolved into her bringing business tactics from auto factories to help the startup, like overhauling their business contracts.
She peppered the team with questions about the science until some of her friends told her she should quit automotive and work full-time in biotech.
“I was like, no one’s gonna take me seriously. I’ve never studied biology. I studied poli sci and women’s studies,” she recalls.
But she made the leap anyway, taking a massive pay cut from her well-paying six-figure job to what amounted to intern-level pay. She learned about startups, raised money, and worked her way up to director of operations. The company went public, giving her a healthy enough payout to buy a house, she said.
Flushed with success, she launched a biotech startup of her own, a contract research lab.
She raised money, then made all the classic first-founder mistakes. “I hired people too quickly. I opened up my lab,” she said.
Two years in, her startup burned through its funds, and she knew she had to shutter it. It broke her heart and her bank account. She even lost her house, she recalled.
But she had gained a stellar reputation in Michigan’s tight-knit startup community and Brown recalls that VCs told her, “You’re a good founder anyway.” Several said they’d be open to funding her next idea.
Knowing she would soon be available for a new venture, Deangelo began sending her scientific material on circular RNA. He had an idea for how to use it with AI drug discovery.
“He started sending me, literally every morning at 5:30 … five to 10 articles,” she remembers. “I hadn’t even shut the other company down all the way.”
But she studied up and grew convinced this idea could work. They founded CircNova in May 2023.
“I went into it very cautiously, throwing just a few things at the wall. What can I do with the $15,000 grant to get it started?”
That first expenditure developed the startup’s first process and another $25,000 from a National Science Foundation grant led to the first patent application.
She began to split her time between Michigan and Boston, near her customers and wish-list customers like Moderna and Pfizer.
As for betting on Brown again, VCs like Nia Batts, a general partner at Union Heritage Ventures, had no problem with it.
“We are no stranger to the resilience that is needed when you engage in the journey of entrepreneurship,” Batts said, adding that she knew she wanted to back this new venture “the moment” she met Brown and heard about the idea.
This $3.3 million seed round was led by diversity-focused VC South Loop Ventures and includes investment from Dug Song, Union Heritage, Michigan Rise, Invest Detroit, Kalamazoo Forward Ventures, and SPARK Capital.
Keep reading the article on Tech Crunch
Five of the most important fintech VCs investing heavily in the sector
Global investing in fintech startups is starting to see an uptick. Just this week, KPMG issued its Pulse of Fintech report for the second half of 2024. In the fourth quarter of 2024, investment climbed to $25.9 billion from $18 billion in the third quarter, according to KPMG.
Granted, this is not the enthusiasm of years past, especially the wild days of 2021. But judging by our inboxes, there is no shortage of venture capitalists still betting big on the space.
Below is a list of just a few VCs who remain bullish on fintech.
About the VC: Infinity Ventures is a three-year-old early-stage venture firm dedicated to investing in fintech startups globally founded by Jeremy Jonker, Jay Ganatra, and Mario Ruiz.
They left PayPal Ventures in May 2021 and closed on their first $158 million fund in October 2021. In October 2024, the firm raised a $184 million Fund II, bringing total assets under management to more than $350 million.
Average check size: Depends on stage: $1-2 million pre-seed; $2-4 million seed; and $5-10 million Series A.
Notable investments: Rainforest, Pagos, Mendel.
Most recent big investment: SimpleClosure, a business-closure platform.
The firm’s focus is on B2B fintech and commerce enablement.
About the VC: Founded in 2019 by NerdWallet co-founder Jake Gibson and Sheel Mohnot, Better Tomorrow Ventures leads rounds in pre-seed and seed-stage fintech companies globally.
It has $225 million in assets under management.
Average check size: Ranges from $500,000 to $4 million.
Notable investments: Unit, Relay, Coast, Mendel, Charlie.
Most recent major investment: Basis (BTV led seed, Khosla led A).
If pitching Better Tomorrow Ventures, Mohnot previously told us: “Find a way to get in front of us that is exciting! We respond to cold emails all the time — here’s an example of a cold email that worked.”
About the VC: Founded in 2022 by “This Week in Fintech” newsletter author Nik Milanović, The Fintech Fund invests in fintech companies globally at the pre-seed and seed stages.
Last September, the firm closed its second $10 million fund.
Average check size: Ranges from $200,000 to $400,000.
Notable investments: Rainforest, Unit, Cascading AI, Ansa.
Most recent major investment: WiseLayer.
If founders are looking for hands-on investors, they’ll find it in The Fintech Fund, according to Milanović.
“There’s a lot of ETFs that will write large checks,” he told TechCrunch previously. “But our goal is to really bring together this whole community — and that’s newsletter readers, investors in the fund, our angel syndicate — so that when the founder gets a check out of The Fintech Fund, it’s not just money but also a ton of consulting or referrals to new hires and to new customers.”
About the VC: Atlanta-based TTV Capital invests in early-stage companies with a focus on traditional fintech, fintech-enabled businesses and the “future of fintech.”
Its assets under management total over $750 million.
Average check size: Ranges from $2 million to $8 million.
Major investments: Green Dot, Bill.com, Greenlight.
Most recent major investments: Charlie, Payabli.
Partner Lizzie (Guynn) Hartley previously told TechCrunch that when it comes to pitching, she prefers to take initial pitch meetings as video calls.
“Before I talk to a founder on a call, they should be able to clearly articulate the problem they are solving. The ability to distill this down into a digestible and understandable statement is very helpful. I appreciate when founders can walk through the customer workflow end-to-end and prove the customer benefit. This helps us build conviction in a customer’s willingness to pay for a new tool or software,” she said.
About the VC: The firm has been around for well over a decade, exclusively investing in companies building financial technology at the pre-seed to Series A stages. It currently has $4 billion in assets under management.
In 2023, QED Investors announced that it had raised $925 million across two new funds to back fintech startups globally — a $650 million early-stage fund and a $275 million growth-stage fund.
Managing partner and co-founder Nigel Morris told TechCrunch at the time of the firm’s last fund close that QED planned to make about 40 investments out of that early-stage fund.
Average check size: $15 million to $20 million.
Major investments: Credit Karma (QED was the first institutional money into the company), Creditas, Nubank, SoFi.
Most recent major investment: Kin Insurance, One Card, Moniepoint.
QED is focused on embedded finance, cross-border payments and wealth management, as well as on AI. Roughly half of QED’s portfolio is in the U.S., followed by LatAm, Europe, Southeast Asia and Africa. It’s also looking into fintech opportunities in the Middle East. One of its more recent investments was in Japan, its first in the country.
Want more fintech news in your inbox? Sign up for TechCrunch Fintech here.
Want to reach out with a tip? Email me at [email protected] or send me a message on Signal at 408.204.3036. You can also send a note to the whole TechCrunch crew at [email protected]. For more secure communications, click here to contact us, which includes SecureDrop and links to encrypted messaging apps.
Keep reading the article on Tech Crunch
Elon Musk Rants Incoherently in Sunglasses While X Shadowbans His Pleading Baby Mama
February 20, 2025 was a big day for the world’s richest man and shadow President Elon Musk. He called an astronaut “fully retarded” during an online fight. The Wall Street Journal reported that his lawyers have been threatening companies with Federal regulatory problems if they don’t spend money advertising on X. And he capped off the day by ignoring the pleas of Grimes, mother of three of his children, as he took the stage at the Conservative Political Action Conference to mumble incoherently while wearing sunglasses and waving around a chainsaw.
To hear Musk and Newsmax host Rob Schmitt tell it, the world’s richest man is now a hero. His poll numbers aren’t trending well, but whatever. Conservative Americans, especially those who attend CPAC, aren’t living in the same reality as the rest of the country. Musk is in his own special echo chamber that’s even more removed from the public. Wealth insulates you from your fellow man, and Musk is the richest man who has ever existed. He’s rich enough to purchase a popular social media platform and turn it into a hall of hollow praise and Nazi memes.
Grimes—the mother of three of Musk’s children, including the one he carries around with him for photo ops—took to X to plead with him yesterday. “Plz respond about our child’s medical crisis,” she said in a post on X. “I am sorry to do this publicly but it is no longer acceptable to ignore this situation. This requires immediate attention.”
She went on at length in several posts. She was replying to something Musk had posted unrelated to her and said Musk had stopped replying to her texts, emails, and phone calls. Newly revealed Musk mother Ashley St. Clair went through something similar earlier this week. Both women tried, and failed, to get Musk’s attention by replying to him on the social media site he owns. The result was the apparent “shadowbanning” of Grimes on the platform.
Musk was pretty busy all day getting into fights with astronauts and constantly posting. Later in the day was his big interview at CPAC. That’s the kind of thing a billionaire needs to focus on. Besides, the medical emergencies of your biological children can be so boring. Musk just wants to have a little fun.
Musk took the stage at CPAC, shook hands with disgraced meme-coin pumping rug-pulling Argentine president Javier Gerardo Milei and picked up a glittering chrome-plated chainsaw. Musk started the interview by waving the chainsaw around the stage and hooting for the audience.
Musk wore sunglasses, a Texas-style belt buckle, a black-on-black MAGA hat, and a gold chain. He didn’t take off the sunglasses for the entire interview and his body-mic picked up the flopping and rubbing of the gold chain as Musk nervously twitched and gesticulated during the conversation.
“I am become meme. There’s living the dream. And there’s living the meme,” Musk said to the audience after he sat down. “DOGE started out as a meme. Think about it. And now it’s real. Isn’t that crazy?”
Then he talked about how much fun he’s having and how everything is so very funny right now. “The left wanted to make comedy illegal. Y’know. You can’t make fun of anything. Like, comedy sucks. Nothing’s funny. You can’t make fun of anything. It’s like, legalize comedy! Yeah! Legalize comedy,” Musk said.
Schmitt asked how Musk found himself on the right side of the political spectrum. The billionaire blanked, like a computer program stuck in a recursive loop. Schmitt prompted him by asking if the left had gone crazy and forced him out.
“Y’know, the whole cancel culture stuff. Freedom of speech,” Musk said. “Infringe upon people’s personal freedoms. They just want state control. State control of what you say. They want to take away your guns.” Musk then rambled about how, now that he’s looked at the various government departments, he doesn’t even believe there’s a real left wing in the country. “I’m not sure how much of the left is even real…it doesn’t have popular support.”
Seventy-five million people voted for Kamala Harris.
Schmitt and Musk did a rapid-fire question-and-answer session about a bunch of conservative bête noires, talking points, and conspiracy theories.
Is it true that DOGE and Trump are going to send Americans $5,000 checks? “It sounds like that’s something we’re gonna do,” Musk said.
On assassinations, Musk would like to know more. “Why do we still know nothing about that guy in Butler? What’s going on? Kash is going to get to the bottom of it!” We know a lot about Thomas Matthew Crooks, the man who took a shot at Trump. It’s been reported largely in the mainstream media that Musk decries.
“We’re fighting the Matrix big time here,” Musk said.
Trump has talked a lot recently about the concern that Fort Knox, which is supposed to hold America’s gold reserve, has no gold. Schmitt asked Musk if he thought the gold was there.
“I don’t know….we wanna go see it and make sure someone didn’t spray paint some lead or something, y’know?” Musk said. “Like, is this real gold? Bite the bar, y’know? Honestly, part of this is just like, let’s have some fun. And like, I said, all this gold at Fort Knox. It’s your gold. It’s the public’s gold, so like, I think you have a right to see it.”
Musk is governing by meme. He just wants to have fun. Slashing tens of thousands of jobs from the federal bureaucracy, accessing the private data of American citizens, and peddling bizarre conspiracy theories to a clapping audience is all just fun and games. “It’s very easy to take advantage of the federal government,” Musk said, with no self awareness whatsoever.
To see Musk out, Schmitt asked a tough question. “How much do you sleep?”
Musk hunched over in his chair, lowered his sunglasses, and darted his eyes back and forth over the crowd. They laughed. Musk didn’t answer.
“Paint us a picture of inside the mind of a genius,” Schmitt said.
“Uh,” Musk said. He made a face at the crowd. They laughed.
“My mind is a storm,” Musk said. “So. It’s a storm. But. But. I mean maybe let me tell you something, I mean, you can ask the question, but I think it’s worth elaborating on something. Which is. I grew up in South Africa and my morality was informed by America. I read comic books. I played Dungeons and Dragons. I watched American TV shows. It seemed like America cared about being the good guys. About doing the right thing. And that’s actually pretty unusual, by the way. So I was like, yeah. You wanna be on the side of good. You wanna care about what’s right. And um. And uh. That’s uh. That’s what I believe in.”
“I gave you a tough one at the end,” Schmitt said.
Musk stared into space. “So, uh, yeah.”
On X, someone pointed out to Grimes that she’d probably been shadowbanned and that it was hard for anyone to see her begging Musk for help with their children.
“lol” Grimes replied.
Report: AI coding assistants aren’t a panacea
As they gain in popularity, AI coding assistants such as GitHub Copilot may appear to be boosting productivity. But in reality, they could be causing overall code quality to decline.
That’s the top-line finding from a new report released by software engineering platform GitClear, which analyzed 211 million code lines from 2020 to 2024. According to GitClear’s analysis, there was a remarkable decline in code reuse last year — a potential cause for concern, given that code reuse is a common practice to help build redundant systems.
Several recent surveys have shown that AI coding assistants tend to produce mixed results.
One from software vendor Harness found the majority of devs spend more time debugging AI-generated code and security vulnerabilities compared to human-written contributions. A Google report, meanwhile, found that AI can quicken code reviews and benefit documentation, but at the cost of delivery stability.
Keep reading the article on Tech Crunch
Apply to Speak at TechCrunch Sessions: AI before the deadline
AI Innovators, seize your moment! Have insights that could inspire 1,200 AI founders, investors, and enthusiasts eager to advance the future of AI?
Take center stage, influence the AI conversation, and exchange ideas at TechCrunch Sessions: AI on June 5 at UC Berkeley’s Zellerbach Hall.
We’re gathering top AI visionaries from the startup world to lead compelling sessions and interactive roundtables. Join us in helping entrepreneurs, founders, and innovators navigate the evolving world of AI.
Lead a breakout session at TC Sessions: AI
This is your moment to delve deep into pressing AI topics. Organize a team of up to four presenters (including a moderator) to guide a 50-minute session featuring a presentation, panel discussion, and audience Q&A to ignite meaningful conversation.
Click the “Apply to Speak” button and submit your topic on this event page. Whether your focus is startups, investments, infrastructure, or emerging AI tools, TC Sessions: AI is the prime stage to share your expertise.
Once your topic is submitted, our audience will vote on it, selecting the sessions they want to experience live at the event.
Don’t wait — the application deadline is March 7.
Breakout speaker perks
It’s more than just branding — get the full TC Sessions: AI experience! As a breakout speaker, you’ll enjoy increased visibility while also benefiting from all the perks of an attendee. Gain access to exclusive AI main-stage discussions, breakouts, and valuable 1:1 or small-group networking opportunities.
Additionally, TechCrunch will help amplify your brand with:
- Event agenda listings (web and app).
- A mention in a featured TechCrunch.com article.
- Social media promotion.
Inspire, educate, and lead! Play a pivotal role in shaping the AI ecosystem and solidify your reputation as a trusted expert in the field. Don’t miss your chance to speak! TC Sessions: AI is on June 5, but the application deadline for content submissions is March 7. Apply now before it’s too late!
Keep reading the article on Tech Crunch
Apple pulls iCloud end-to-end encryption feature for UK users after government demanded backdoor
Apple confirmed Friday that it “can no longer” offer a security feature that allows users in the United Kingdom to encrypt their iCloud data.
In a statement provided to TechCrunch, Apple spokesperson Fred Sainz said the company’s Advanced Data Protection feature will no longer be available to new users and current U.K. users “will eventually need to disable this security feature.”
“We are gravely disappointed that the protections provided by ADP will not be available to our customers in the U.K. given the continuing rise of data breaches and other threats to customer privacy,” the company said.
“Enhancing the security of cloud storage with end-to-end encryption is more urgent than ever before,” the statement said.
The announcement comes after the U.K. government reportedly ordered Apple earlier this year to build a backdoor that would allow British authorities “blanket” access to users’ data stored on Apple’s cloud servers, even if it is end-to-end encrypted. This request, seen as unprecedented in a modern democracy, alarmed privacy and security experts, who argued that if the British government prevailed, the demand would set a precedent for authoritarian countries to follow.
Apple offers users the option to turn on end-to-end encrypted iCloud backups through Advanced Data Protection. This feature effectively makes it impossible for anyone, including Apple and government authorities, to view data stored in iCloud by users’ who have opted-in.
A spokesperson for the U.K. Home Office did not immediately respond to TechCrunch’s request for comment.
Apple did not immediately say how the process of disabling ADP will work for users who had already turned it on before Friday.
Apple said that some types of data, including health data, messages stored in iCloud, and payment information, which are end-to-end encrypted by default for all users, will not be affected by this change, and will remain encrypted for everyone. But U.K. users will not be able to opt-in to use end-to-end encryption for the other types of data, such as photos, notes, backups, and other data, which were encrypted under ADP.
For those who already have ADP enabled, Apple said it will give customers more guidance soon, as well as a period of time to disable the feature to keep using iCloud.
ADP is unaffected for users outside of the United Kingdom, Apple said, and end-to-end encrypted communication services like FaceTime and iMessage are not affected, either.
“As we have said many times before, we have never built a backdoor or master key to any of our products or services and we never will,” Apple said, linking to its prior statements.
BBC News reported that ADP stopped being an option for new users starting at 3 p.m. U.K. time on Friday. TechCrunch has also confirmed that ADP is no longer an option for new users in the United Kingdom.
Since the rise of encryption in the mid-1990s, governments worldwide have argued that this data-scrambling technology would allow criminals and terrorists to break the law while evading law enforcement. Over the years, authorities have always found a way, from accessing unencrypted backups to using spyware, to access data directly on people’s devices.
“If you are not in the U.K., you should turn on ADP now,” said Matthew Green, a cryptography expert and teacher at Johns Hopkins University, wrote on X in response to the news.
“The more people who use it, the harder it will be to shut off this way,” said Green.
Clarified the forms of data protected under Advanced Data Protection in the ninth paragraph.
Keep reading the article on Tech Crunch
DeepSeek to open-source parts of online services code
Chinese AI lab DeepSeek plans to open-source portions of its online services’ code as part of an “open source week” event next week.
DeepSeek will open-source five code repositories that have been “documented, deployed and battle-tested in production,” the company said in a post on X on Thursday.
Code repositories are storage locations for software development assets, and typically contain source code as well as configuration files and project documentation.
“As part of the open-source community, we believe that every line shared becomes collective momentum that accelerates the journey,” the company wrote. “Daily unlocks are coming soon. No ivory towers — just pure garage-energy and community-driven innovation.”
DeepSeek, which has a history of making its AI models openly available under permissive licenses, has lit a fire under AI incumbents like OpenAI. In recent social media posts, OpenAI CEO Sam Altman admitted DeepSeek has lessened OpenAI’s technological lead, and said that OpenAI would consider open-sourcing more of its technology in the future.
Keep reading the article on Tech Crunch
A huge trove of leaked Black Basta chat logs expose the ransomware gang’s key members and victims
A trove of chat logs allegedly belonging to the Black Basta ransomware group has leaked online, exposing key members of the prolific Russia-linked gang.
The chatlogs, which include over 200,000 messages spanning from September 18, 2023, to September 28, 2024, were shared with threat intelligence company Prodaft by a leaker. The cybersecurity firm says the leak comes amid “internal conflict” within the Black Basta group after some members allegedly failed to provide its victims with functional decryption tools despite paying a ransom demand.
It’s not yet known if the leaker, who uses the alias “ExploitWhispers” on Telegram, was a member of the Black Basta gang.
Black Basta is a prolific Russian-language ransomware gang, which the U.S. government has linked to hundreds of attacks on critical infrastructure and global businesses, whose publicly known victims include U.S. healthcare organization Ascension, U.K. utility company Southern Water, and British outsourcing giant Capita. The leaked chat logs give a never-before-seen look inside the ransomware gang, including some of its unreported targets.
According to a post on X by Prodaft, the leaker said that the hackers “crossed the line” by targeting Russian domestic banks.
“So we are dedicated to uncovering the truth and investigating Black Basta’s next steps,” the leaker wrote.
Targeted victims, exploits, and a teenage hacker
TechCrunch obtained a copy of the hackers’ chat logs from Prodaft, which contain details about key members of the ransomware gang.
These members include “YY” (Black Basta’s main administrator); “Lapa” (another of Black Basta’s key leaders); “Cortes” (a hacker linked to the Qakbot botnet); and “Trump” (also known as “AA” and “GG”).
The hacker “Trump” is believed to be an alias used by Oleg Nefedovaka, who Prodaft researchers describe as “the group’s main boss.” The researchers linked Nefedovaka to the now-defunct Conti ransomware group, which shut down soon after its internal chat logs leaked following the gang declaring its support for Russia’s full-scale invasion of Ukraine in 2022.
The leaked Black Basta chat logs also quote one member as saying they are 17-years-old, TechCrunch has seen.
By our count, the leaked chats contain 380 unique links related to company information hosted on Zoominfo, a data broker that collects and sells access to businesses and their employees, which the chatlogs show the hackers used to research the companies they targeted. The links also give some indication of the number of organizations targeted by the gang during the 12-month period.
The chat logs also reveal unprecedented insights into the group’s operations. The messages include details on Black Basta’s victims, copies of phishing templates used in their cyberattacks, some of the exploits used by the gang, cryptocurrency addresses associated with ransom payments, and details about ransom demands and victims’ negotiations with hacked organizations.
We also found chat logs of the hackers discussing a TechCrunch article about ongoing Qakbot activity, despite an earlier FBI takedown operation aimed at knocking the notorious botnet offline.
TechCrunch also found chat logs that named several previously unknown targeted organizations. This includes the failed U.S. automotive giant Fisker; healthtech provider Cerner Corp, which is now owned by Oracle; and U.K.-based travel firm Hotelplan. It is not yet known if the companies were breached, and none of the companies responded to TechCrunch’s inquiries.
The chat logs appear to show the gang’s efforts in exploiting security bugs in enterprise network devices, such as routers and firewalls that sit on the perimeter of a company’s network and act as digital gatekeepers.
The hackers boasted their ability to exploit vulnerabilities in Citrix remote access products to break into at least two company networks. The gang also talked about exploiting vulnerabilities in Ivanti, Palo Alto Networks and Fortinet software to carry out cyberattacks.
A conversation between Black Basta members also suggests that some of the group were worried about being investigated by Russian authorities in response to geopolitical pressures. While Russia has long been a safe haven for ransomware gangs, Black Basta was also concerned about actions brought by the U.S. government.
Messages sent after the group’s breach of Ascension’s systems warned that the FBI and CISA are “100% obliged” to get involved and could lead to the agencies “taking a tough stance on Black Basta.”
Black Basta’s dark web leak site, which it uses to publicly extort victims into paying the gang a ransom demand, was offline at the time of publication.
Keep reading the article on Tech Crunch
TechCrunch Disrupt 2025: Lowest prices of the year end in 7 days
You read that headline correctly! The best deals for TechCrunch Disrupt 2025 tickets are about to end in just 7 days. Save up to $1,130 on individual passes and up to 30% on group tickets. Don’t wait — these offers end on February 28 at 11:59 p.m. PT.
Join us in celebrating 20 years of TechCrunch Disrupt from October 27-29 at Moscone West in San Francisco. Connect with 10,000+ tech leaders, dive into 250+ sessions, and gain valuable insights from 200+ experts. Plus, experience the legendary Startup Battlefield 200 and next-level AI insights.
Register now to secure the biggest Disrupt ticket savings of 2025.
What awaits you at Disrupt 2025
AI deep dives: Explore AI topics spanning healthcare, transportation, SaaS, policy, defense, hardware, and beyond.
Expert insights: Gain wisdom from 200+ leaders covering business scaling, leadership, and key industries like space tech, fintech, IPO, and SaaS to fuel your growth.
Interactive sessions: Engage in live Q&As or deepen your knowledge in expert-led roundtables and breakout discussions.
Startup Pitch Battle: Watch startups, hand-selected by TechCrunch, compete in Startup Battlefield 200 for the $100,000 equity-free prize and Disrupt Cup. Learn from world-renowned VC judges about building a viable business. Previous winners include Dropbox, Fitbit, Trello, and Cloudflare.
Are you that startup or do you know a startup that should compete? Add the startup’s information to the Startup Battlefield waitlist to be the first to know when applications open.
Build valuable connections: Connect with the leaders shaping tech’s future. Whether networking with investors, seeking mentors, or finding new business partners, Disrupt is where it all thrives.
Be part of Disrupt 2025 — get your tickets at the best prices before the rates go up after February 28.
Fueling 20 years of innovation
For 20 years, TechCrunch Disrupt has been the hub for pioneering founders, visionary tech leaders, and key investors to drive the future of entrepreneurship. It’s the place where investors connect with the innovators reshaping tomorrow’s tech landscape.
Here are some of the groundbreaking leaders who’ve appeared on the Disrupt stage:
- Alex Pall and Drew Taggart from The Chainsmokers, Co-founders and Partners, MANTIS Venture Capital
- Ashton Kutcher, Co-founder, Sound Ventures
- Assaf Rappaport, Co-founder and CEO, Wiz
- Bridgit Mendler, CEO, Northwood Space
- Colin Kaepernick, Founder and CEO, Lumi
- Denise Dresser, CEO, Slack
- Erin and Sara Foster, Co-founders and General Partners, Oversubscribed Ventures
- Mary Barra, CEO, General Motors
- Matt Mullenweg, Co-founder, WordPress and CEO, Automattic
- Peter Beck, Founder and CEO, Rocket Lab
- Serena Williams, Founding and Managing Partner, Serena Ventures
- Shaquille O’Neal, Entrepreneur and Philanthropist
- Vinod Khosla, Founder, Khosla Ventures
- RJ Scaringe, CEO, Rivian
Lock in the lowest prices of the year for Disrupt 2025 tickets
Don’t let up to $1,130 in savings slip away! Grab yours today before this deal ends on February 28 at 11:59 p.m. PT. Register now to lock in the best rates of the year.
Is your company interested in sponsoring or exhibiting at TechCrunch Disrupt 2025? Contact our sponsorship sales team by filling out this form.
Keep reading the article on Tech Crunch