Eighteen months after their merger fell apart over who would control whom, Honda and Nissan have agreed to share the software inside their cars.

The two signed a joint development agreement on Monday covering electronic control units, an in-vehicle operating system, middleware and vehicle control software, with the first cars arriving from fiscal 2029.

The agreement is narrower than a merger and, in engineering terms, more intimate. Standardising core ECUs means agreeing on the computers that run the brakes, the steering, the battery and the cabin, and on the interfaces between them.

Both companies framed the decision as a matter of pace and cost. Faster development cycles and more efficient investment were the stated reasons, after what they described as extensive studies of where collaboration in the software domain might work.

Neither disclosed what it will cost or save. No supplier was named, no executive was quoted by name, and the specific ECUs to be standardised were not identified.

Fiscal 2029 is the part worth sitting with. That is roughly three years from now for architecture work that has to be locked long before the first vehicle reaches a showroom, which tells you how far ahead this kind of decision is made and how little room there is to change it afterwards.

The pieces being standardised are the ones customers never see and never forgive. An operating system that stutters, an over-the-air update that fails, a control unit that disagrees with another: these are the faults that define whether a car feels modern.

The problem they are addressing is real and not specific to Japan. A modern car runs on tens of millions of lines of code across dozens of controllers, and every manufacturer is discovering that writing all of it alone is expensive in a way that engines never were.

Scale is the only available answer. Arm has been arguing that the industry needs common foundations rather than dozens of bespoke stacks, which is precisely what two carmakers sharing an operating system amounts to.

The competitive pressure is coming from a specific direction. BYD has built its own 4nm driving chip and is fitting LiDAR to a $10,000 car, which is a level of vertical integration neither Japanese company can match alone.

Chinese brands have been taking share in Europe at a record rate, and the software experience inside the vehicle is a large part of why. Buyers comparing a Chinese EV with a Japanese one in 2029 will be comparing operating systems as much as drivetrains.

What makes this agreement notable is the history behind it. Merger talks collapsed in February 2025 after Honda proposed making Nissan a wholly owned subsidiary, an arrangement Nissan would not accept, and the two walked away.

Sharing an operating system without sharing ownership is a considerably harder thing to sustain. Every future disagreement about a feature, a release schedule or a security patch now has to be resolved between two independent boards with different product cycles and different problems.

There is a precedent both will know well. Shared platforms have worked in this industry for decades, but they involved metal and tooling with fixed specifications, whereas software has to keep being decided long after the vehicle ships.

Nissan has the more urgent problems of the two, having spent the period since the merger collapse restructuring. Honda is the larger and steadier partner, which is the imbalance that sank the merger and has not gone anywhere.

The pair have said they will keep exploring collaboration elsewhere, naming carbon neutrality and the elimination of traffic fatalities as areas of shared interest. Those are aspirations rather than agreements, and the distinction is doing work in the announcement.

What exists today is a signed document, a defined technical scope, and a date in 2029. Whether it holds together until then is a question about two organisations rather than about software.

You’ll need your old passcode or Apple Account password.

How to fix an iPhone Unavailable message

Tatiana Diuvbanova/Shutterstock

Nearly every company that’s trying to deploy artificial intelligence runs into the same problem: it’s hard to integrate the tech into everyday operations. Industrial heavyweight Caterpillar has spent decades dealing with a version of that problem in the physical world, and now it’s using its experience to deploy AI.

Caterpillar’s push into the autonomous space started with mining, where labor shortages and hazardous conditions can make automation particularly useful. Today, it sells automated haul trucks, drilling, underground loaders, dozers, remote-controlled construction equipment, and more. It also offers a software command center, fleet management, and even remote terrain intelligence as part of its autonomous toolkit.

“Now we’re in this super exciting time where we can take all of that learning from mining and bring it into much more dynamic environments, jobsites, quarries, and construction sites,” the company’s CTO, Jaime Mineart, told TechCrunch on during a fireside chat at the Ai4 conference in Las Vegas earlier this month.

The industrial giant is now applying AI more broadly, including in tools used by technicians and its own employees. One example is the Cat AI Assistant, which lets field technicians standing next to a machine use voice commands to pull up repair procedures, troubleshoot potential problems, and identify parts that may be needed before beginning a repair. Mineart said the tool is now being used by customers, operators and technicians.

The assistant draws on Caterpillar’s proprietary data, which spans information generated by its connected machines. Mineart said Caterpillar has about 1.6 million connected assets globally and more than 16 petabytes of structured data.

The company is also using AI to power software for scanning sites and generating digital twins in manufacturing to analyze operations, she said. And like nearly every other company, Caterpillar is using AI across its enterprise operations, as well as for software development. “We use AI agents to modernize legacy code, generate and test new software, and identify defects earlier,” Mineart said.

But Mineart is quick to point out that building the technology is only part of the challenge, as deploying an autonomous machine is not the same as transforming a site to use AI. Companies also have to rethink how people work alongside the technology and how existing processes need to change.

“The hard part about autonomy and about physical AI is incorporating that technology into the customer jobsite and into the workflows,” she said.

Mineart said the company leans on experienced operators to help train AI systems, leveraging institutional knowledge built over decades. And as machines become more autonomous, some operators may shift from controlling a single machine to overseeing multiple machines from a remote command center.

That transition, however, is creating a new challenge for Caterpillar: training its 118,000 employees. Mineart said the company plans to spend $100 million over the next five years to train its workforce in AI, autonomy and robotics.

That investment is likely being put towards helping the company make the most of the broader boom in AI infrastructure, which is already helping its top-line. Caterpillar’s quarterly revenue reached an all-time high of $20.5 billion in the second quarter, helped by strong demand for power-generation equipment used in data centers. Its power-generation division saw sales spike 72% to $3.10 billion, and CEO Joe Creed said that “no one is slowing down” when it comes to demand for cloud computing and generative AI infrastructure.

Krafton has unveiled five upcoming games as part of its plans to create global gaming franchises. The company announced them during its third consecutive Krafton Media Day at Gamescom 2026 in Germany. Its new lineup includes titles from different genres, along with a new take on the PUBG franchise.

Krafton introduced five upcoming titles at gamescom 2026, covering different genres and offering new gameplay ideas.

• NO LAW (Neon Giant): Claes af Buren, CEO of Neon Giant, presented the open-world FPS RPG NO LAW. The game takes place in Port Desire, a city where law and order have collapsed. Players’ decisions can change the challenges, opportunities, and stories they encounter. Neon Giant also uses the latest Unreal technologies to give players more freedom to shape their experience.
• Project ZETA (NIRVANANA): Namseok Kim, CEO of NIRVANANA, showcased Project ZETA, its Multi-team Tactical Arena game. Unlike traditional MOBAs, the title does not use a fixed lane-based structure. Four teams, each with three players, enter objective-focused battles from the start. NIRVANANA plans to refine the game through global testing and player feedback.
• Age Twisters (Piccolo Studio): Piccolo Studio Co-founders and Directors Jordi Ministral and Alexis Corominas introduced Age Twisters. The two-player adventure follows a grandfather and granddaughter through four different stages of life. Players will see their age differences reflected through unique abilities and changing conversations. The game continues Piccolo Studio’s focus on creating experiences with an emotional connection.
• TARAE: The Unbound (Boundary): Boundary CEO Inyoung Koo presented TARAE: The Unbound, an Eastern dark fantasy action RPG. The game draws inspiration from the concepts of reincarnation and karma in Buddhist thought. East Asian mythology and aesthetics also influence its world, character classes, combat, and story. This gives the game a setting and identity beyond the Western medieval style often seen in dark fantasy games.
• PUBG: DED.NET (PUBG STUDIOS): PUBG STUDIOS Creative Director Dave Curd introduced PUBG: DED.NET as a new direction for the franchise. The game keeps the survival and competitive elements associated with PUBG: BATTLEGROUNDS but moves away from its battle royale format. It takes place in Cascadia in 1996, inspired by the Pacific Northwest during the 1990s. A roguelite structure lets player decisions influence each run and what happens next.

All five games Krafton will release will be on display at gamescom 2026 in Cologne, Germany, from August 26 to 30 at Koelnmesse, Hall 9. This showcase reflects the company’s policy of creating global franchises and accommodating the creative concepts of its developers.

The company was originally added to a list of organisations thought to be aiding China’s military in January 2025, and that status was upheld in June of this year.

Chinese chipmaker ChangXin Memory Technologies (CXMT) is suing the US Department of Defense to seek its removal from a US government blacklist.

The company was originally added to a list of organisations thought to be aiding China’s military in January 2025, and that status was upheld in June of this year.

Established in 2016 in Hefei, CXMT creates the dynamic random-access memory (DRAM) chips needed to power mobile phones, PCs, tablets, servers, and a range of consumer products and applications.

“CXMT designs, produces and sells its DRAM chips for civilian and commercial use, ​not for military use,” it reportedly said in its court filings.

US defence secretary Pete Hegseth and two of his staff are among the named co-defendants in the suit, filed last Friday (28 August) in a US district court for the District of Columbia.

“CXMT is not a military company and has no affiliation with the Chinese military,” the company said in a statement after the filing, according to Bloomberg.

“Since its initial designation in January 2025, CXMT has continuously suffered reputational and commercial harm. It is pursuing this action to protect its reputation and business interests.”

CXMT is on a list of “entities identified as Chinese military companies operating in the United States”, according to the US defence department, which are categorised as such under a ‘section 1260H’ designation.

Designated businesses are also said to be “engaged in providing commercial services, manufacturing, producing or exporting … and operate directly or indirectly” in the US. The list also features companies such as Tencent, Huawei, BYD and Alibaba.

Earlier this month, CXMT surpassed Tencent in market capitalisation to become China’s most valuable company at around $520bn.

Last week, CXMT reported first-half revenues for 2026 of around $22bn – up more than 870pc year-on-year – fuelled by the global memory crunch rooted in the AI-driven wave of demand for chips.

US curbs on semiconductor exports to China over recent years have pushed China to focus on its own chip capabilities. Earlier this year, China reportedly began restricting overseas travel for its top AI professionals in its escalating battle with the US over technological dominance.

Last week, a US judge blocked the Pentagon’s blacklisting of products made by US AI giant Anthropic being used for government purposes, ruling that its actions violated the company’s rights.

Don’t miss out on the knowledge you need to succeed. Sign up for the Daily Brief, Silicon Republic’s digest of need-to-know sci-tech news.

The takeaway: Nvidia’s latest 13F reveals a $63.4 billion portfolio of US stocks, and it is very lopsided. Intel and SpaceX account for roughly 80% of it. But the filing’s most interesting detail is what didn’t happen: seven of the eight positions were untouched last quarter, but the portfolio more than tripled in value anyway.

Nvidia’s earnings this week gave the market the growth story it wanted. But a separate SEC filing gives us an additional perspective on what Nvidia is also doing with all those billions.

Form 13F is a quarterly disclosure document any institutional manager holding more than $100 million in US equities has to file. But Nvidia is a chip designer, not a hedge fund, which is precisely what makes the filing worth reviewing.

As of June 30, the company held $63.44 billion in public stock across eight positions as shown below. That’s every public US stock Nvidia owns, but that does not mean these are the only investments the company has on its books. Not even close…

* Intel stock was valued in June 30, but the stock has moved downwards since: the stake is now worth roughly $19.2 billion as of late August.

** Nvidia’s real stake in Nebius is undercounted in the 13F filing. Nvidia holds about 22.2 million Nebius shares, or 9.3% of the company (worth approximately $4.68 billion). Much of those shares are tied to a warrant that can’t be exercised until September 11, 2026. Only the common shares show up on the table above.

Six of the eight positions have a disclosed cost. Together they came to roughly $22.4 billion and were worth about $63.1 billion at the end of June – a potential paper gain of $40 billion on companies Nvidia mostly bought into because it needed them to succeed, not because it expected the trade itself to work. Three months earlier that portfolio was worth $18.4 billion.

SpaceX’s stake and everything the filing leaves out

Several reports we’ve seen of this filing describe SpaceX as a “new position.” But it’s not. It’s a newly visible one. Nvidia put $10 billion into xAI in January as part of a $20 billion round. In February, SpaceX absorbed xAI in an all-stock deal. Then when SpaceX listed on the Nasdaq in the largest IPO in history, Nvidia’s private stake became 122.76 million Class A shares of a public company, and therefore something the SEC required it to disclose.

That’s also a useful reminder that the 13F is not everything Nvidia owns. A 13F only captures US-listed equities and says nothing about private companies, which is where most of Nvidia’s money has actually gone.

Credit: App Economy Insights

The filing does not show the ~$30 billion stake in OpenAI, the investment in Anthropic, or any of the other private AI labs on the books. It also omits positions Nvidia has publicly announced in companies such as Lumentum and Marvell, along with the $2 billion warrant it wrote for Nebius – which is why Nebius appears here as a rounding error despite being a significant recipient of Nvidia capital.

Nvidia has reportedly deployed north of $100 billion into AI companies over two years. The 13F captures only a slice of that.

A portfolio or a bill of materials?

As for the picks Nvidia is making, they are all intertwined. Coherent makes the optics that move data between GPUs inside a rack. Nokia builds the network gear that moves it between buildings. Synopsys makes the EDA software Nvidia’s own engineers use to design chips. Intel is a potential second-source foundry and a CPU partner. CoreWeave and Nebius are neoclouds that exist to rent out Nvidia hardware by the hour.

Even the smallest position fits the pattern. Generate Biomedicines (0.02% of the portfolio), uses AI to design drugs. So every single holding is either a supplier Nvidia needs to move faster or a customer Nvidia needs to keep buying. That is the whole argument behind the “circular financing” complaint that’s been going around for the last year or so.

Source: Bloomberg

Meanwhile, Nvidia was ready to respond to those criticisms. CFO Colette Kress acknowledged the label directly, said Nvidia “sees it differently,” and framed the stakes as investments in once-in-a-generation companies where the downside is limited. CEO Jensen Huang was less diplomatic on CNBC. He said this is the first generation of startups that needs tens of billions of dollars before it can turn a profit, and somebody has to write those checks. However, the risk is that those companies’ rise and fall are on the same AI demand curve as Nvidia’s chip sales.

The Intel stake shows how fast that can move. Worth roughly $30 billion on June 30, it had slipped to about $19 billion weeks later after an Intel stock offering diluted it.

A blast from the not-so-distant past (July 2020):
Nvidia is now worth more than Intel

Nebius runs the same problem in reverse. The 13F lists it at $329 million, which makes it look like an afterthought. But Nvidia actually controls 9.3% of the company – most of it not yet common stock, but sit in a warrant from Nvidia’s $2 billion investment that can’t be exercised until September 11 – that investment is currently valuated at around $4.68 billion.

The generous view is that this is simply what a platform company does when the technology transition is real and its customers are capital-starved. Nvidia is buying the parts of the supply chain that would otherwise slow it down, and it is doing so at a moment when it has more cash than almost anyone in history.

The less generous view is that Nvidia has wired its balance sheet so that a slowdown in AI spending hits twice, in the same quarter, in the same direction. Chip revenue falls, and the value of the companies buying those chips falls with it.

However, both can also be true at once. What the 13F does is make the wiring visible, and there’s a lot more of it that this filing still doesn’t show.

When George Arison took over Grindr in 2022, he inherited a company that had been bounced from Chinese ownership to a forced divestiture to a private-equity rescue — a business that was printing money but also had no real product or business strategy. Four years, a SPAC listing, and a controversial return-to-office mandate later, Grindr has become convincing as a growth story. Revenue is on pace to roughly triple, from $195 million in 2022 to a guided $540 million-plus this year, with adjusted EBITDA margins holding above 40%.

That growth has come almost entirely from getting existing customers to pay more versus dramatically growing its user base. In the second quarter of this year, it had 1.4 million paying users, or 9% of its user base, but average revenue per user has risen considerably since 2022, and Arison is very focused on where the next leg of growth comes from. Part of that plan includes turning Grindr into a “gayborhood in your pocket” — a platform that handles not just dating and hookups but healthcare (from ED medication to HIV prevention to, eventually, connecting users with gay doctors) and travel (helping users find community wherever they land). It’s the same “everything app” instinct driving much of consumer tech right now.

But that’s not the only thing Arison is throwing against the wall; later this year, Grindr is rolling out a subscription that it’s betting the market will support, a far pricier “EDGE” tier that has already raised hackles from some on the internet (“literally who’s paying for this,” and “we need 2012 grindr back”).

Arison, who previously founded and ran Shift Technologies, the online used-car marketplace he also took public via a SPAC, in 2020, is also aggressively seeking out press to argue that institutional investors continue discounting Grindr’s stock simply because it’s a gay dating app. In a Zoom call with TechCrunch on Thursday, he cited an investor who once showed him a financial model with a literal “Grindr discount” line item, knocking 25% off a fair-value estimate.

Not everyone thinks the discount is justified. Morgan Stanley, Goldman Sachs, and Raymond James have all raised their price targets on the stock this year, and Morgan Stanley upgraded it to “overweight” in July, citing the EDGE tier and Grindr’s telehealth push — part of a run-up that’s seen the stock climb roughly a third over the past six months. Still, that discount isn’t Arison’s imagination — the stock trades at roughly 11 times 2027 EBITDA, about a 35% discount to peers — even if it’s not clear why.

Arison — amiable, with a hint of his native Georgia still in his voice — was game to talk about all of it in a Q&A that has been edited for clarity and length.

You came in to run Grindr in 2022 when it was pretty adrift. What was the first thing you wanted to prioritize?

Three things, really. First, company building. A lot of people had been hired during the peak-COVID era when expectations for in-office productivity were low, and average output was maybe three to four hours of real work a day. We brought people back to the office two days a week in the summer of 2023 — a decision that got a lot of press and a lot of anger. We went down to about 70 employees as a result. Today only around 25 people who were here before I arrived still work at Grindr, and we run lean: 175 U.S. employees plus a team in Colombia, doing $540 million in guided revenue this year.

Second was driving revenue growth by shipping product people would actually pay for — that’s how pay conversion went from under 6% to over 9%, and ARPU nearly doubled. Third was setting a long-term vision: the “gayborhood in your pocket” idea, and everything else — healthcare, travel — that flows from it.

You’ve said Grindr’s engineering culture is underrated. How small is the team actually running this?

About 94 or 95 people across all technical roles. A large-tech-company CEO once told me AI would let me do with far fewer people what I thought I’d need 300 to 350 for — he was right. We’re doing roughly 350 people’s worth of work with about 100. Something like 80% of our code is now AI-written, and we’ve seen a 2.5x increase in engineering productivity over the past year.

Grindr tested an AI-powered premium tier, EDGE, at a price that in Canada worked out to roughly $350–375 a month in U.S. dollars — a number that got mocked online as more expensive than just dating someone. What actually happened there?

We haven’t released EDGE — we’re testing it, and some people have access as part of that test. It sits above our existing XTRA ($23.99) and Unlimited ($44.99) tiers. We’re not selling AI itself; we’re selling features derived from it — using what we know about a user’s behavior and intent, with consent, to make much better matches than a sparse profile ever could. Retention on these features so far is higher than anything we’ve had before.

The pricing people quoted was one test point among several — we ran a range to understand elasticity, not a final price. EDGE goes live toward the end of this year or early next, and we’ll know where it lands by then. We think of it like a Tesla Model X or S: a premium flagship now, with the underlying capability rolling down to the broader product over time.

On the matching itself — you’ve described using AI to suggest partners outside someone’s home city, because gay-dating pools even in places like San Francisco are small. What evidence do you have that those long-distance matches lead anywhere?

Even in San Francisco, where the share of the gay population is higher than almost anywhere else in the country, you’re talking about maybe 50,000 to 60,000 gay people total. That’s not a big pool to be fishing in if you’re trying to find a partner, and it’s a big reason dating is so hard for gay men generally. So yes — what if AI could break down the geographic constraint entirely, and surface someone in St. Louis who actually matches what you want, based on real behavior rather than what people say in a profile?

As for whether it leads anywhere real: we don’t track people’s relationships after the fact — that’s going too far. What we do know is that Grindr is where most gay men say they meet each other for relationships, and that younger cohorts want something different than older ones did: about 50% of gay men under 35 say they want a long-term monogamous relationship, and 25% say they want children — numbers that would have been unthinkable for my own generation. When you ask people why they’re not in a relationship, the answer is usually that they have a hard time finding a partner. I can’t promise this solves that. But it’s worth trying something new, because what’s been done so far clearly hasn’t solved it.

There are so many ways to expand your addressable market beyond the dating app itself. The healthcare stuff is really interesting — can you tell us more about what you’re doing there, and whether you’re building these products yourselves or acting as the interface to outside providers?

We started with cash-pay products through a line we call Woodwork — ED medications, GLP-1s, peptides, and so on — because cash-pay was the simplest place to get going, and we just launched an AI bot that handles the whole transaction inside the app rather than sending people out to Woodwork.com. The second bucket is HIV prevention and treatment: we’ve committed to giving 10 million people direct access to information on where to get PrEP, both in the U.S., where we already offer that in our in-app health center, and internationally.

The third bucket — actual clinical care, like connecting people to a gay doctor through telehealth — is very much long-term. That’s not something we’re building today, but I do think there’s a world, a decade out, where healthcare is a bigger revenue stream for Grindr than what we do today.

Right now, though, non-subscription revenue — ads and everything else, including healthcare — is a small fraction of the business.

Subscriptions are about 83% of revenue today, down from around 86% in 2022, even though subscription revenue itself has grown enormously — that tells you how much bigger the base is overall. The newer businesses are genuinely small right now. What I want is a company that, 10 years from now, has a strong subscription business, a strong advertising business, a real healthcare business, and a real travel business alongside it. Today those last two are early.

You’ve said investors still apply what one called a “Grindr discount” to the stock because of what the company is. But the stock is up sharply over the past six months, Morgan Stanley just upgraded it, and it trades at a premium multiple to Match Group. Isn’t the market telling you the discount is gone?

I hope we’re being treated as a growth company at this point — we’ve grown revenue more than 25% for 16 straight quarters as CEO, so there’s reason to expect that. The stigma conversation is a real one; we’ve had a consulting firm decline to work with us over reputational concerns, and a bank refuse our money during the Silicon Valley Bank crisis, even as other major banks like Goldman and Morgan Stanley have been strong partners. I think a lot of that is really about Grindr being a gay dating product rather than dating being controversial per se — nobody says that about Tinder, which literally has a “free tonight” button on its homepage. But I’d agree the market’s read on us has clearly improved.

The MSI Claw 8 EX AI+ is a high-octane handheld games console that can run PC and Xbox games from a wide range of game stores. With the Intel Arc G3 Extreme chipset — featuring the Intel Arc B390 GPU and 32GB RAM — it’s easily the most powerful portable console I’ve ever (quite literally) gotten my hands on. But there’s a pretty huge catch to that performance: a price tag that would buy you two Steam Deck OLEDs or a mid-price gaming laptop.

Let’s start with the superficial stuff: the MSI Claw 8 EX AI+ looks gorgeous. Its screen is an extensive eight inches across, offers wonderfully rich colors and a silky-smooth 120Hz refresh rate. Combined with its discreet RGB lighting around the asymmetrically placed thumbsticks and Void Purple colorway, it has a unique look that definitely helps it stand out from the crowd.

I filed a request with McDonald’s earlier this month to access all of the personal data the fast food company collected about me, and I received a stunning 515-page report a few days later that detailed my app interactions in granular detail and predicted I would never stop eating there.

Under the California Consumer Privacy Act, I have the legal right to request access to information from large companies that collect personal data. So I was curious what others might have on me, and I spent the next week filing more than 100 requests.

The CCPA went into effect in 2020, and three of its key provisions are the right to opt out of the selling of personal information, the right to delete that info, and the right to request a copy for yourself.

I focused solely on the latter—access requests—to better understand what data is being collected. Most companies must list two ways for you to file. These are often via a web form, phone number, or email address, as designated in their privacy policy. After you submit a request, companies can take 45 days to complete it.

My experience placing these data access requests was incredibly time-consuming, from finding the right filing methods to verifying my identity multiple times. Most exasperating during this process were the companies that either responded to my access requests with messages concerning the deletion of information, which I explicitly said not to do, or refused to process the request through a method listed in their privacy policy.

Consumer advocates I spoke with were upset with how these requests were handled. “That’s crazy,” said Ben Winters, director of AI and privacy at the Consumer Federation of America. “That’s not an acceptable status quo.” Winters sees these examples as exhibiting the weaknesses of policy frameworks that rely on companies to act responsibly and in good faith.

In accordance with WIRED’s policies, I am disclosing that I used generative AI to draft bureaucratic emails and update my tracking spreadsheet as part of this report. I wrote the body of this article mainly by hand in my scratch notebook.

One of the first errors came from Crunchbase, known for its database about tech startups. I emailed my access request to its privacy address on August 17. My message laid out the rights I wanted to exercise and included a direct request not to erase anything: “I am not requesting deletion at this time. Please do not treat this as a deletion request.” I received a reply two days later from a Crunchbase support representative.

“Thanks so much for your patience. Your account has been permanently deleted from Crunchbase. Please let me know if you need anything else!” the message read in full.

I followed up via email almost immediately, reiterating that I wanted data access, not data deletion. “Your Crunchbase user account was deleted. Other data located on Crunchbase was not deleted,” read the follow-up support response explaining what happened. If I wanted to have a Crunchbase account, I would have to reregister.

When I reached out to Crunchbase for comment, a spokesperson blamed the mistake on a “processing error” and said that the company would proceed with my original access request as filed. The spokesperson also claimed the misclassified response came from “a person on our customer success team” and not a generative AI tool.

My interactions with BeenVerified, a searchable database that gathers public records, also encapsulate my friction-filled experience placing these access requests.

I emailed BeenVerified’s dedicated CCPA compliance address on the morning of August 19. It laid out that I was a California resident placing an access request, not a deletion request. You’ll never guess what happened next.