Oura, the smart ring maker, with offices in San Francisco and Finland, is reportedly planning to raise up $3 billion as soon as next month in a U.S. IPO that values the more than 900-person company at north of $16 billion, according to Bloomberg. The outlet adds that investors are expected to sell a major chunk of stock in the offering.

A $16 billion valuation would mark an enormous jump from the $10.9 billion valuation Oura was assigned last September, when it closed an $875 million Series E round backed by Fidelity, ICONIQ, Whale Rock, and Atreides, joining earlier backers like Dexcom, The Chernin Group, Forerunner Ventures, Coatue, and Temasek.

The wearables space has gotten crowded fast.

Samsung launched its own ring, the Galaxy Ring, two years ago. But Oura’s most direct rival may be the fitness band maker Whoop, which has undergone its own reinvention. Once a performance tool for elite athletes and young men, Whoop has spent the past year courting a much broader audience, layering in hormone tracking and blood-panel testing for perimenopause and thyroid health. The move drove its own valuation up to $10 billion back in March. Oura has followed a similar path, evolving from its starting niche of biohacking CEOs into a more mainstream sleep-and-recovery brand.

Oura announced in May that it had filed confidentially for an IPO.

Not all of Oura’s recent attention has been positive. A proposed class action lawsuit filed last week in San Francisco is accusing Oura of misleading consumers about the accuracy of its sleep tracking features. Among other complaints in the filing, Oura is accused of not just claiming “to measure a heartbeat or a temperature, but the exact stage of sleep the wearer is in—which in reality requires electrodes in the scalp and sensors on the eyes, as only a hospital or other clinical setting can do.”

In a statement to TechCrunch sent over the weekend, a spokesperson pushed back against claims in that suit, saying that: “While Oura Ring is not a medical device or a substitute for a clinical sleep study,” said the company, “Oura’s sleep staging has been validated and compared favorably in multiple studies against polysomnography, the gold standard . . .Multiple third-party, independent studies support our claims of accuracy and we have transparently reported on the mechanisms and measures that inform Oura’s sleep staging.”

The company disputes the allegations and “will defend against them in the appropriate legal forum,” added its representative.

Zillow and Redfin have reached a settlement with the Federal Trade Commission (FTC) and five states, ending a legal fight over a 2025 partnership that the FTC claimed hurt competition in the rental-listing market. The settlement was announced on Monday, just as the case was scheduled to head toward trial this morning.

The case stems from a deal announced last year in which Redfin agreed to display Zillow’s rental listings on its websites rather than compete directly with Zillow for rental advertisers. The arrangement could have kept Redfin out of the rental advertising business for as long as nine years. Redfin owns Rent.com and ApartmentGuide.com, two major rental-listing platforms.

According to the FTC and attorneys general from Arizona, Connecticut, New York, Virginia, and Washington, Zillow agreed to pay Redfin $100 million to keep Redfin from competing with Zillow.

The companies defended the partnership as a way to give renters access to a larger pool of listings. The FTC, however, argued that Zillow was paying one of its largest competitors to stop competing, potentially allowing the company to charge higher prices and provide less favorable terms to property managers. It could have also reduced the quality of rental listings available to consumers.

Under the proposed settlement, Redfin will be required to reenter the rental advertising business. The order also removes restrictions that previously limited Redfin’s ability to compete independently for property-management customers.

The settlement doesn’t completely end the relationship between the two companies. Redfin can continue displaying Zillow’s rental listings, but it will once again be able to compete for its own customers. Redfin will be able to sell advertising, display listings from its own clients, and pursue new rental customers without being required to share sensitive business information with Zillow.

The Zillow-Redfin case comes just months after the DOJ’s settlement with Ticketmaster, another antitrust case involving allegations that a dominant company used its power to suppress competition. However, 26 of the 30 state attorneys general who initially sued Live Nation alongside the DOJ chose to continue pursuing the case and won their lawsuit in April.

You don’t need a keynote session to make noise at TechCrunch Disrupt 2026. Sometimes the best way to meet investors, customers, and partners is by exhibiting directly on the Expo Hall floor at San Francisco’s Moscone West from October 13-15.

That’s exactly what our Exhibit Program offers. Showcase your startup and get the added bonus of additional tickets for Disrupt to bring even more members of your team and make the most of this exciting opportunity.

All the perks of Disrupt’s Exhibit Tables

Our exhibit tables start at $12,500, and when you break down all the included benefits, the returns speak for themselves. Here’s what’s included:

A 6’ x 30″ table in the Expo Hall, which is the highest-trafficked part of Disrupt, where more than 10,000 founders, VCs, and tech leaders come looking for partnerships, investments, and networking.

We provide the table, linen, chairs, and signage so you can focus on showcasing your startup — not logistics.

10 Disrupt passes, including five all-access Exhibitor-Partner passes and five Expo+ passes, so your whole team can work the floor, not just staff a booth. Those passes cover nearly $4,000 of the exhibition price by themselves!

Lead generation through the Disrupt 2026 mobile app, so every conversation turns into a follow-up, not a lost business card.

Access to the Disrupt 2026 press list and Silver Tier sponsor branding across the site, app, and signage.

Joining fellow exhibitors is the fastest, easiest way for a startup at any stage to get in front of the exact room that matters, from investors sourcing deals, to operators evaluating tools, to founders scouting partners.

There is one catch, though: Tables are limited. This offer ends when the last exhibit table sells out, or on September 25 at 11:59 p.m. PT, whichever comes first.

Book your exhibit table now. Or, if you’re not ready to exhibit this year, get a Disrupt 2026 ticket while we’re running a limited-time $300 discount and start planning your presence for next year!

Alabama’s attorney general announced Monday that it sent a subpoena to OpenAI as part of an investigation into the company’s alleged “complete lack of oversight and adequate safeguards” in the Hugging Face incident.

The investigation comes weeks after OpenAI admitted that one of its unreleased and guardrail-free cybersecurity models had escaped an isolated environment, connected to the internet, and hacked AI dataset platform Hugging Face. As Reuters first reported, Hugging Face was only one of four victims of what was supposed to be “an internal evaluation” of a model with “maximal cyber capabilities,” as OpenAI put it.

The press release announcing the subpoena sent by the state’s attorney general Steve Marshall said that the state was seeking to understand if OpenAI’s “inability or unwillingness to ensure the safety of its products” violated the state’s consumer protection laws.

When reached by TechCrunch for comment, OpenAI spokesperson Nate Evans provided the statement: “The Hugging Face incident marked an important moment for AI safety and we are conducting a thorough review along with external advisors. Once the review is complete, we will share a technical report with relevant government authorities and publish our findings publicly.”

Earlier this month, Marshall, along with the attorneys general of 14 other states, including Florida, Missouri, Pennsylvania, and Texas, sent a letter to OpenAI’s CEO Sam Altman, requesting that he and his company preserve all records related to the Hugging Face incident. The letter also asked OpenAI to “immediately cease and desist” from any internal cybersecurity evaluations.

In the wake of the Hugging Face incident, and several other incidents disclosed by Anthropic, the U.K.’s AI Security Institute, Meta, and workers at AI companies — including executives and technical leaders — signed an open letter called “Pacing the Frontier,” which called for developing AI capabilities slowly and more responsibly. The letter also called for the U.S. government to support an “international effort to develop the technical and governance tools needed to deliberately pace the frontier of automated AI development.”

Updated with comment from OpenAI.

In Brief

Over the weekend, Amazon significantly raised the prices of its hardware devices, jacking them up by as much as 60% in some cases.

The price explosion impacted Fire TVs, Echos, Kindles, and Eeros. One egregious example that’s been cited is that of the Echo Dot, one of Amazon’s cheapest smart speakers, the price of which jumped 60% overnight, from $49.99 to $79.99. Price-tracking sites like CamelCamelCamel show the stark uptick, which has previously hovered much lower.

When reached for comment by TechCrunch, Amazon shared the following: “The consumer electronics industry is facing significant increases in memory and storage component costs. After absorbing these increases for as long as we could, we recently adjusted pricing across our product lines.”

The company also said that it would continue to offer occasional promotions to customers over the course of the next year.

Amazon’s price hikes reinforce the reality that it’s a terrible time to buy hardware. The global memory shortage fueled by the AI boom (also known as “RAMmageddon“) is driving up the cost of producing devices of all stripes. As a result, companies are increasingly passing on the costs to their customers.

In a sign of the times, Apple recently hiked its prices, and has sought to offset the steep new rates by introducing a device-leasing program, thus allowing users to pay off their devices over time.

The RAM shortage is expected to continue throughout 2027 before prices potentially crest and stabilize in 2028.

The AI boom has made it possible for a significantly wider range of people to dip into development, create code, and launch software for the first time. There are pros and cons to this, even for established developers who are being pushed to use it to produce work faster.

Replit sits at the center of this evolution, and when its CEO and co-founder, Amjad Masad, takes center stage at TechCrunch Disrupt 2026, he’ll speak about the future of programming and the wide-ranging implications of a world in which ideas can be easily turned into products.

Join the audience to watch our wide-ranging discussion with Masad at San Francisco’s Moscone West from October 13–15. You’ll also catch the rest of our excellent Disrupt lineup across six stages that address the most pressing questions facing the startup community, including, most critically, how do you build a sustainable company in the AI era?

Get your tickets now during our special promotional pricing extension — get up to $300 off your ticket until 11:59 p.m. PT Tuesday, August 25.

Why Replit is joining the Disrupt Stage

Replit’s rise has been explosive, mirroring the broader ascent of AI and the companies furthering its proliferation. Though Masad founded the company a decade ago, the past 18 months have marked the biggest changes, with the company’s current run-rate tracking toward a billion dollars annually, compared with just $2.8 million in reported revenue in 2024.

But this conversation is about much more than Replit’s revenue. Masad’s conversation will dive into his perspective about the future of programming, entrepreneurship, and how Replit ensures that its service is doing more than helping users prototype their ideas — it helps fully flesh them out, too.

Investors are convinced of Replit’s value to both nontechnical users and professional coders. They assigned the company a $9 billion valuation earlier this year, a mere six months after the company was assigned a $3 billion valuation.

Of course, the question is whether it can keep that momentum going — or whether its growth in 2026 will look quaint a year from now because it’s growing even faster.

If you’re interested in building within the AI era, you can count on many sessions, in addition to Masad’s, across our other stages:

Builders Stage, which will be chock-full of VCs and corporate development pros who’ll dive deep, providing the kinds of insights every founder needs to have.

AI Stage, where we’ll examine the presumed SaaSpocalypse, career opportunities, emerging security challenges, and more.

Smart Systems Stage, for those working on the vital infrastructure underpinning this era and for those who need to learn more about it.

Real World AI Stage, where we’ll explore the sundry ways that new technologies are beginning to impact the physical world around us.

To hear it first and in person, join the more than 10,000 founders, investors, and technologists who are heading to San Francisco for Disrupt 2026 on October 13–15. Grab your ticket today to make sure you get the best available price.

Learn more about Disrupt 2026

Check out Disrupt’s headline speakers

Everything founders should know about Disrupt

Get the best hotel deals ahead of Disrupt

How to host your own Side Event at Disrupt

Take part in Disrupt and learn how to exhibit your startup

In Brief

President Donald Trump bought as much as $50,000 worth of SpaceX shares on June 23, according to a financial disclosure first reported by Reuters, two weeks after the record-setting IPO of Elon Musk’s company.

It’s not clear what price Trump paid for the shares, but by that point they had fallen from their highs of over $200. SpaceX shares were trading in the mid-$150 range on June 23. At the end of trading on Monday, shares closed at the IPO price of $135, possibly putting the president’s stake underwater.

Trump and Musk are close, despite a brief falling out last summer that involved the businessman accusing the president of withholding the Department of Justice’s files on Jeffrey Epstein because of how often Trump’s name appears in them. SpaceX has been hoovering up an increasing amount of government contracts and benefiting from the Trump administration’s deregulatory stance, according to a recent Wall Street Journal analysis.

White House spokesman Davis Ingle told Reuters that the president’s stock portfolio is managed by third-party financial institutions and replicate “recognized indexes, such as the Schwab ​1000.” SpaceX lobbied popular indexes to change their rules to allow for faster inclusion ahead of its IPO, which means many people likely own some of the company’s stock even if they don’t know it.

Everyone is buzzing about Instinct, an AI personal assistant still in private access, not only for its incredible capabilities, but also for its potential privacy concerns. The agent, a veritable taskmaster, has been praised as feeling “like magic” and being one of the “most exciting launches” since OpenClaw. However, some testers have also raised concerns about the AI agent’s security model and its worrisome terms of service.

To be fair, Instinct is still in private testing for now, so these concerns haven’t yet scaled to the wider public at this time.

Created by a small team led by former Sierra research scientist Noah Shinn, San Francisco-based Instinct is operated by Spear Street Technology, per its terms and California business filings. It’s currently operating in stealth, per PitchBook.

The AI agent itself works by connecting to your applications and devices, including your email, messaging apps, calendar, and your device’s audio, location, screen, and more. You can text the agent or call it via text message or WhatsApp, asking it to perform various tasks for you, like booking your appointments and reservations, scheduling a ride to the airport, cleaning up your inbox, organizing important information, handling your shopping, finding you cheap flights, and much more.

Though Instinct is described by testers as outperforming their expectations, some have also raised concerns about the company’s approach to customer privacy and security. This raises a timely question: Are the trade-offs of giving AI this level of access and autonomy actually worth it?

For instance, several people are circulating screenshots from the company’s Terms of Service, which grant Instinct a broad “perpetual and irrevocable” license to “access, use, host, cache, store, reproduce, transmit, display, publish, distribute, and modify” any of the user’s materials, including for training its AI models. The terms also detail how Instinct can receive information from users’ devices, including screen captures, cursor movements, and keyboard inputs.

The terms also allow Instinct to enter into “agreements, commitments, or transactions” on users’ behalf, which would be binding.

let’s pause for a sec and talk about https://t.co/19wLOjliYq



VCs, notable people, successful people hyping it up all week. They all just giving away their emails and all their personal data, emails, messages:



“sub-licensable, worldwide, perpetual and irrevocable license to… pic.twitter.com/MWb3LKnvQp — Mike Khristo (@MikeKhristo) August 22, 2026

From a #cyberhealth perspective, Instinct is a hard no.



On the positive side, their policy is 100% forthcoming.



However, the access they require comes with responsibility I would not bestow on any company.



For those who desire speed or the "current thing" if you want to be… https://t.co/HdXO1k18u3 pic.twitter.com/VczQw9P0Sa — Jeremy Banon (@jeremybanon) August 21, 2026

One early adopter, Peter Yang, pointed out that Instinct would not delete his Gmail records when asked. (The team later fixed the problem by adding a tool for deleting external data in its settings, he said.)

Hey Instinct, it's not cool to index and retain my emails without my permission and not let me delete them from your records?



I can't recommend this to anyone until this is figured out. Thanks @clairevo for spotting this. pic.twitter.com/J2qhYVqwxN — Peter Yang (@petergyang) August 21, 2026

Another person, Claire Vo, found that Instinct was still summarizing their inbox after disconnecting its access. When she asked Instinct what happened, the bot confirmed that the emails were stored in plain text for later searches.

what a thrill!



i disconnected ai bot instinct from google at 11 AM and got a summary of my emails at 2 PM 😵



here's what happened, and what i've learned 👇 pic.twitter.com/LMGGLXcrD5 — claire vo 🖤 (@clairevo) August 21, 2026

Many others questioned the security model, too. One tester was a bit worried when they found that Instinct was able to pull a sign-up code from their email inbox to complete a particular task — in their case, booking a table at a restaurant via Resy. And Hello Patient co-founder Alex Cohen wrote that once he found out how easily Instinct could be phished, he deleted his account:

For additional context, I don’t think we’re at the point where it’s safe to give AI read/write access to your inbox.



I wanted to see how easily Instinct could be phished, so I created a brand new Gmail account and emailed my real personal account with instructions for Instinct… https://t.co/Mh1W3xJQov pic.twitter.com/V5inRtK9Zj — Alex Cohen (@anothercohen) August 22, 2026

In addition, Moxxie Ventures founder Katie Jacobs Stanton shared that Instinct broke her trust when it sent an email on her behalf without first checking with her.

“We’re trading privacy and control for hyper-personalized AI tools (AI notetakers, personalized AI agents, etc), often without fully understanding the trade,” she remarked on X, summarizing the dilemma posed personal AI agents. “The more powerful these agents become, the more trust matters. Every successful action earns a little more trust. One unauthorized action can reset that trust to zero,” she said.

Instinct is an amazing product. I've been using it for mostly personal needs and a little bit of work.



Last night, it was a little naughty and sent an innocuous email on my behalf without checking with me first. I told it that it had broken my trust and disconnected my email.… — Katie Jacobs Stanton (@KatieS) August 22, 2026

Michael Mignano, the founder of Anchor, which was acquired by Spotify, and now a GP at Union Square Ventures, noted that products like Instinct are going to “change modern security norms for consumers,” adding that “people will increasingly hand over passwords to 3p [third-party] apps, unaware of how or what they are storing for them.”

Interest in Instinct and personal AI has been growing since the arrival of OpenClaw, a personal AI assistant that became popular for its powerful capabilities, leading its founder to join OpenAI to help work on the next generation of personal agents. Another messaging-based assistant, Poke, also just exited to Cognition.

Amid the criticism, Instinct’s team hasn’t yet responded to anyone’s concerns or complaints on X, preferring to keep a low profile. (The bot itself identifies Luca Borletti, also formerly of Sierra, as being involved with the company, but that has not been confirmed.) TechCrunch has heard from multiple investors that Kleiner Perkins and Conviction have invested in the startup and those rounds have now closed.

I've been using Instinct every day for the last week now. It’s been awesome for travel booking, rebookings, restaurant reservations, email follow-ups, CRM management, even working on our data room for LPs. I've tried Hermes, OpenClaw, Tasklet, GrokBot but Instinct takes the cake… — Jesse Middleton (@srcasm) August 22, 2026

Requests for comment sent to both the startup’s main email address and Shinn directly have not yet been returned.

We heard the feedback, and we’ve extended the window for the lowest prices for TechCrunch Disrupt 2026 a bit further, with the current prices now ending at 11:59 pm PT Tuesday, August 25. Once that window closes, your chance to save up to $300 on your pass, or get an even larger discount on group pricing, will come to an end.

For everyone who was looking for an extension, your chance is now! Start getting ready to join the rest of the startup community gathering in San Francisco from October 13-15 at Moscone West to learn from speakers from OpenAI, Anthropic, Replit, Amazon, Databricks and more.

If Disrupt has been on your radar, now is the time to lock in your pass before tomorrow’s deadline.

TechCrunch Disrupt: Where the startup ecosystem moves forward

Each year, TechCrunch Disrupt brings together more than 10,000 founders, investors, and startup community members, driving innovation forward. And for our event this year, you’ll find a keen focus on building and developing in the AI era across all of our stages of programming. If you want to be on the cutting edge running into 2027, this is the event for you.

And you don’t have to be a founder to get value out of Disrupt. Whether you’re raising capital, scouting investments, hiring talent, launching a startup, or building strategic partnerships, Disrupt puts you in the middle of the conversations shaping what’s next and can help you make the connections that propel your next year of growth.

Here’s just a glimpse of what you gain by attending:

Actionable insights from founders, operators, and VCs actively building and investing in today’s market. Our agenda features a wide range of topics, curated to focus on insights that lead to action.

from founders, operators, and VCs actively building and investing in today’s market. Our agenda features a wide range of topics, curated to focus on insights that lead to action. Direct access to investors looking for their next portfolio company and founders seeking the right partners. Plus, you can get AI-powered networking opportunities through our app or take deeper dives through our many Side Events.

looking for their next portfolio company and founders seeking the right partners. Plus, you can get AI-powered networking opportunities through our app or take deeper dives through our many Side Events. Early visibility into emerging technologies and startups before they break into the mainstream through our Exhibit Hall.

into emerging technologies and startups before they break into the mainstream through our Exhibit Hall. High-value connections that lead to funding, partnerships, customers, and career opportunities. Disrupt is about growth, both for startups and yourself!

What’s new at TechCrunch Disrupt 2026

Keeping that focus on building companies in the AI era in mind, we have a great slate of new stages and programming to inspire and educate:

Real World AI Stage: AI is moving beyond the screen and into the physical world. Explore how robotics, autonomous systems, manufacturing, healthcare, and defense are turning AI breakthroughs into real-world products and businesses.

AI is moving beyond the screen and into the physical world. Explore how robotics, autonomous systems, manufacturing, healthcare, and defense are turning AI breakthroughs into real-world products and businesses. Smart Money Stage: Follow the money. From fintech to stablecoins to payments, embedded finance, and AI-driven financial services, this stage explores how technology is reshaping the movement of capital.

Follow the money. From fintech to stablecoins to payments, embedded finance, and AI-driven financial services, this stage explores how technology is reshaping the movement of capital. Smart Systems Stage: Every AI breakthrough depends on the infrastructure behind it. Discover the innovations in chips, compute, energy, networking, and data centers that will determine the next generation of technology companies.

Our Disrupt, AI, and Builders Stages are also returning this year, and we have excellent speakers joining us across them all:

You can explore the rest of our extensive speaker lineup right here!

Don’t miss out on the best TechCrunch Disrupt’s prices!

This is your last chance to save before rates increase, so lock in your pass before Tuesday, August 25 at 11:59 p.m. PT to get the best prices, whether you’re a student looking to get their introduction to the community or an established investor in search of an untapped opportunity.

New York-based General Intuition, the startup building a foundation model that trains generalized AI agents to move through space and time, is in talks to raise funding at a $6 billion pre-money valuation from new investors including Valor Equity Partners, Point72 Ventures, and Seven Seven Six, according to sources familiar with the matter.

Existing investors including Khosla Ventures and General Catalyst are also participating in the round. The fresh funds would come just weeks after General Intuition raised $320 million at a $2.3 billion valuation.

The news confirms earlier reporting on the deal for what is becoming one of the hottest startups tackling physical AI.

CEO Pim de Witte spun out General Intuition last October from his video game clip-sharing platform Medal, using its hundreds of millions of hours of gameplay and “action labels” — records of which buttons a player pressed and when — as an initial dataset.

First sneak peak in the WSJ on how GI’s large action models work! https://t.co/6fuOcu9fkb — Pim de Witte (@PimDeWitte) August 21, 2026

Investor Vinod Khosla recently told TechCrunch that he believes those action labels will be a key part of the “emergence of intuition,” or the ability for a model to truly generalize across tasks that it wasn’t explicitly trained on.

General Intuition is still finalizing the round, and a source close to the deal said that it is oversubscribed as the startup continues to field interest from investors. The company intends to use the funds to focus to improve its general model with a focus on robotic embodiments. That means spending more on compute infrastructure (the company has a partnership with neolab CoreWeave) and hiring more talent.

Valor Equity Partners is known for backing SpaceX. TechCrunch has reached out to confirm its investment in General Intuition, which would be the first AI lab the fund has invested in since SpaceX.



This story is developing. Please check back in for updates.

How much control are you willing to give an LLM over your digital life?

Getting the most value from a model means giving it the keys. For a control freak or the AI-hesitant, it seems like a lot. For Andrew Ambrosino, the lead engineer for OpenAI’s desktop app, it’s the only way to test the future, which is why that app now has access to, and control over, his inbox, his Slack account, his phone, apps like Notion and Figma, and more.

“If I’m asking it to write a document, is there a possibility that it’s going to pull from a private DM on that subject and not know that it’s not supposed to share some info? Yes,” Ambrosino told TechCrunch. “I’ll do it for the job. I will take the personal hit here and there if I have to. And I haven’t had to.”

Ambrosino works on OpenAI’s biggest bet, ChatGPT Work, which was released last month and is available on the company’s lowest subscription tier, for $20 a month. The product is intended to allow white-collar workers to field AI agents — hooking LLMs up to the digital workflows used by accountants, investors, doctors, and everyone else whose day-to-day is dominated by their computer.

OpenAI’s marketing copy puts the goal succinctly: A world where “where [artificial] intelligence goes beyond answering questions to helping everyone turn their biggest ideas into reality.”

For software developers, that shift is already happening, but it’s been slow to spread to other departments. ChatGPT Work is a modified version of the company’s Codex coding tool. It’s meant to give non-engineers a version of the same functionality that software engineers already get from agents: an AI tool that doesn’t just answer questions, but completes multistep projects on its own.

“In this new factor, ChatGPT can actually do entire, very complicated tasks for you all autonomously in a way that is delightful and safe,” Thibault Sottiaux, who leads OpenAI’s core product work, including Work, told TechCrunch. “It’s the very mission of OpenAI — to bring everyone along.”

Commercially, that matters a lot. Agents that work for longer stretches burn through more tokens, which makes them more lucrative for OpenAI on a per-user basis. Reaching new professions is crucial — not just for OpenAI, but for the industry at large. If coding has proven lucrative territory for AI labs, it’s still a tiny subset of the professional work AI tools need to enable if these companies are to justify their massive investment in training and computation. While labs have been focused on software engineers, vertical-specific competitors like Harvey (for law) and Clay (for sales) have been chasing those customers with a model-agnostic approach, meaning they’ll plug in whichever AI works best at the time.

Industry analysts see this as one of the major challenges facing OpenAI and its competitors. “If the labs cannot rapidly get ahold of the key complementary assets needed to scale AI in the market, value will accrue elsewhere,” Christian Catalini wrote on a16z’s “It’s time to build” blog.

Making the AI apps work for people who aren’t software engineers requires more hand-holding. OpenAI’s non-engineering workforce, like the communications and finance teams, started using Codex “at a time that it was actively hostile to them—asking them about code and showing them, ‘oh, you have an empty diff for this thing,’” Ambrosino said, referring to a technical readout meant for software changes. “So, we started to make it more general purpose between February and now.”

A chart showing Codex usage among OpenAI subscribers, from “The Shift to Agentic AI: Evidence from Codex.” Image Credits:Johnston et. al.

An OpenAI-backed study found that in June, 98% of OpenAI employees were using Codex, but just 17% of organizational subscribers and less than 1% of individual subscribers were using the agentic coding tool. That difference between near total adoption inside the company and negligible adoption outside it is the challenge and opportunity for the company.



“The more value and the more utility that we generate for users, the more they will be willing to also pay for some part of that utility, and that’s how we’ve always seen ChatGPT as well,” Sottiaux said. “You sit there and you’re like, ’of course I want to pay $20 bucks a month for this,’ because the value that you get is so much more.”

How to make AI intuitive

To understand that disconnect, it helps to understand what OpenAI’s engineers are building. Every LLM requires what engineers call a “harness” — the software wrapped around a model that decides what information it sees, which tools it can use, and how it presents its answers back to you.

If you want that model to do stuff — to become an agent — the harness gives it tools and instructions for using them on long-term tasks. For developers, a command-line interface (CLI) that enabled LLMs to code was enough to change the way software was built and deployed. But most people aren’t using CLIs; there’s a reason Windows replaced DOS.

An agentic product that goes beyond software engineering is “going to be something that plays with the messy world of your life and your tools and websites that were built in 1995 and never updated,” Ambrosino told TechCrunch, explaining that the experiences his team is building are vital to expanding access to useful AI.

Consider apps like Claude Code and Codex: They unleashed “vibe coding” by abstracting away all the actual software writing, and letting users just tell the model what they want in a program. Now, OpenAI wants to make functionality found in tools like OpenClaw, which coders use to put LLMs to work, as easy as prompting.

“Without these products in front of the model, experts would know how to get the same results, but you wouldn’t get to a billion people using the thing,” Ambrosino said. That trade-off between what power users need and what mainstream adoption requires plays out in internal debates at OpenAI, where some employees argue that a button is unnecessary if users can just ask the model directly.

“We push back on [that] — because it’s very early,” Ambrosino said. “Discoverability matters in this phase, and at some point we won’t have the button.”

Work has a few more buttons for selecting projects and plug-ins, but it aims for the same magic box interface as other OpenAI products. He compares it to skeuomorphism, the fading practice of making digital tools look like the physical objects they replaced, like a calculator app made to look like a pocket calculator. “That stuff wasn’t just cringe design. That actually helped get people into this [and] make the transition,” Ambrosino said.

OpenAI wouldn’t say how many people used Work versus Codex, but the joint app is used by just 20 million people, compared to more than a billion users the company says are prompting ChatGPT online.

Giving ChatGPT a license to skill

For now, OpenAI is pitching this tool as best suited for routine, data-intensive coordination tasks. Its employees are setting up weekly metrics reports, for example, and making spreadsheets into planning tools.

I’ve spoken to VCs using agents to assemble relevant communications and analysis about companies into investment memos, and ops teams spinning up bespoke dashboards and data visualizations. Sam Altman is using it to plan his vacations. One OpenAI engineer described asking the program to look at a Slack conversation about an engineering problem and “make some charts,” then receiving back a series of insightful plots.

“There is a deluge of information for the average worker or employee of any of these companies, including myself,” Akshay Nathan, who leads the product engineering team at OpenAI, said. “We’re actually quite limited by our ability to parse everything that’s available to us, and then take action on it. That information lives in all these system records tools [like, Salesforce]…the value of ChatGPT is you already have access to this, but now you truly have access to it.”

This, then, could be the digital personal assistant that AI evangelists dream about. As with Claude Cowork or Perplexity AI browsing agent, ChatGPT Work links agents to your existing workspace — email, web browser, a slew of SaaS platforms — and puts that context to work for you.

When the system works, it can be impressive: I asked ChatGPT Work to get my son’s weirdly-formatted preschool calendar out of my email and put it into my Google Calendar, and it did, saving me a lot of repetitive data entry. Hopefully now I won’t forget the school potluck or fail to arrange vacation childcare.

I didn’t trust OpenAI with access to my inbox, source interviews, or story drafts (fear not, AI haters) and wouldn’t let it have access to my bank account, but I believe it would be more useful had I the faith. I tasked it to do financial analysis on publicly traded companies that I cover, and it delivered an auto-updating dashboard of metrics for me; it made a queryable database of space launches, a task I’d previously had to accomplish by writing Python scripts. It also sends me a weekly email about new AI research posted at academic clearinghouses. I’ll keep experimenting with it.

While asking the model for something is intuitive, giving it what it needs to take action isn’t as simple. Setting up the permissions for agents to access, say, a cloud drive was confusing and circular — I tried multiple times to give it just “read” access and received error messages. The model itself wasn’t too helpful, but eventually on the mobile app, a dialog box popped up to tell me that only complete access would make it work.

Many important settings are only available on the web app, so I frequently found myself working in both at the same time. Sometimes ChatGPT Work’s limitations are baffling — link it to your Google calendar and it can create events, but not new calendars. And don’t bother trying to do anything unless the effort level is high, otherwise you’ve got the worst intern you’ve ever worked with.

That’s common advice from AI early adopters, who fear that frustrated newbies will give up. Joe Gershenson, the engineering lead for OpenAI’s harness, admitted that effort settings aren’t intuitive for new users yet — ”there are things that we can do better to help them get the right level of reasoning…” he said, adding, “Watch this space.”

OpenAI faces another important challenge breaking into normie white-collar work: Most workflows aren’t as measurable — or evaluable — as code. Software either works or it doesn’t, and even that distinction reduces the nuance about what makes code good or bad. A good presentation, business strategy, or sales pitch isn’t as easy to evaluate or trace.

“One of the unique challenges with a product like this is just that it can really do anything,” Ambrosino said. When I asked which specific problems the team designs around, and which workflows it targets, the engineers I spoke with demurred, saying that was a question for OpenAI’s research team.

OpenAI later provided an answer, telling TechCrunch that it uses its benchmark GDPval, drawn from 44 occupations and hundreds of knowledge work tests, and supplements that with user feedback. A less official answer is that it comes from OpenAI employees themselves. Said Ambrosino, “We have to always parse out … are we doing the workflow that everybody else will be doing, or are we weird?”

The early adopters of the app itself will create valuable traces with their actual usage — much of the success of coding tools is built on similar data collection — assuming they don’t opt out of making it available for training. (I did).

The rivalry that drove OpenAI’s product design

Despite all the attention on the model interface, OpenAI’s engineers were reluctant to answer a fairly simple question: What sets Codex and ChatGPT Work apart from Claude Cowork, or other competing agentic harnesses intended for a mass user base?



“It’s going to be a really disappointing answer for you, and I’m sorry, but the honest answer is that I really don’t look at the harnesses that they’re building,” Gershenson said in a typical answer. “The Mad Men ‘I don’t think about you at all’ meme comes to mind here.”

Frankly, I don’t believe them, if only based on the extreme similarities between the products’ user interfaces, the need for competitive intelligence at any business, and because the first thing ChatGPT Work asked me to do when I started it up was port over my Claude Cowork data.

It’s understandable if Claude Code is a sensitive topic around the OpenAI offices. Their corporate rival defined the market for AI coding and launched a revolution in how software engineers do their jobs. It’s additionally frustrating because OpenAI had the idea first, but didn’t quite harness it correctly.

When OpenAI first developed Codex as a web app, the engineers got a bit over their skis — or, as Ambrosino puts it, “a bit more AGI-pilled.” In short, they bet on the model being smart enough to handle a task entirely on its own, with minimal user input.

Built shortly afterward, Claude Code was oriented around a back-and-forth conversation with the user. If you gave it a problem, it would survey the possibilities and give you three or four options for proceeding. Once you chose, it would go a little further and then check back again, giving continual updates and leaving less room for the model and harness to make mistakes.

Anthropic’s approach proved more effective, even if it demanded more work from users. “[Our] product was a little ahead of where the model and harness was at the time,” Ambrosino says now.

OpenAI eventually followed suit by adding more opportunities for users to interact with the model. That became the Codex we know today, with desktop and mobile apps. Using download statistics as a proxy for interest in the programs, Claude Code was more in demand until April of this year, but now Codex has taken a slight lead; surveys of enterprise use also suggest OpenAI is catching up.

Part of that lead is getting the product-market fit right, and part comes from complaints about safety restrictions on Anthropic’s models and compute shortages. OpenAI’s steps toward more human-centric harness continue with ChatGPT Work, but the engineers I spoke to insisted the key differentiator is the strength of OpenAI’s latest powerful and cost-effective models.

“The frustrating answer is that a lot of times it is the model, and one thing that we have tried to do really well with this app is fully leverage the model,” Ambrosino said.

What makes a good harness, anyway?

That explanation returns to the “bitter lesson” learned by AI researchers that a better general model is more important than specific domain experience. For true believers, the harness is a temporary crutch, not the moat.

“You could get good results in the short term by adding a whole bunch of extras — if and thens and tools — but like, come on, the next model is going to come out in a couple of months and make that obsolete,” Gershenson told TechCrunch. His team focuses on the simplest ways to expose the model to the tools and context it needs — and no more.

“The goal of good harness engineering is to… be more precise about what information the model really needs to solve your problem, because the models are getting better and better at doing that if you simply let them do their thing,” Gershenson said.

There is an open question, though, if most people or models are ready for that. Ethan Mollick, the Wharton School of Business professor who studies AI tools in the workplace, still sees Claude as more user-friendly, writing that “ChatGPT tends to want to do magic & just do it for you, while Claude does comparisons & shows them, repeatedly asking for input & feedback and doing A & B tests.”

Sottiaux, and perhaps OpenAI at large, disagree, arguing that the conversational nature of the app is better than learning how to use an application. “We definitely see that the world seems to be ready,” he says of the app. ”This is why we’ve had incredible adoption.”

Still, it’s not clear that a model-specific harness is even the right bet for maximizing a model. Comparisons run by companies like Composio and Databricks show that different harness and model combinations deliver different performance on coding benchmarks. Databricks found that Pi, an open source harness published by the software company Earendil, outperformed Codex while using the same GPT 5.5 model. Pi has been used to build projects like OpenClaw and Cloudflare OS.

Pi’s creator, Mario Zechner, says his intentionally minimalist harness is evidence that an AGI-pilled approach can work, at least for software engineers and coding tasks. What it lacks in explicit features, he says, is made up for by its ability to modify itself and build its own interfaces. He sympathizes with the challenge that OpenAI’s engineers face in expanding their user base beyond engineers.

“Everything is coding agent shaped…the reason is that they only have training data for coding agent tasks,” he told TechCrunch. “Say I’m in management, I make a decision today, and the outcome happens months later. You cannot capture that in a simple trace of a user and agent back and forth, so all of these kinds of tasks and anything that you don’t digitize is inaccessible to a model to learn.”

Like other open source providers, he sees the big lab’s effort to push their harnesses as a way to lock-in users; ”They need to own the entire stack; otherwise, they just become a model provider and then need to compete with Chinese models.”

He and other engineers TechCrunch spoke to felt that the insight into token spend and agent behavior in frontier labs’ harnesses is too limited. In a sense, that’s less meaningful to non-technical workers, but as with the coding tools, uptake at the scale OpenAI hopes for will eventually force harder conversations about cost.

For example, messing around on a $20-a-month subscription, I used more than 80 million tokens in four days, which cost $65, according to the model’s analysis (there’s no dashboard in the app). That’s a subsidy of more than 3x the subscription price for four days of casual use alone.

“We are working every day to push the frontier on efficiency,” Sottiaux said, pointing to a recent 80% price cut for users of OpenAI’s Luna model. “If you wake up six months from now, you should be able to do all of the same [tasks] with less spend.”

The other relevant question is whether these apps create a lock-in effect on customers through data retention, or the sheer pain of configuring access to all the plug-ins and their permissions.

Inside OpenAI’s wood-paneled, plant-filled headquarters, which I visited in July, the atmosphere was calm but slightly tense; these are people with a lot to do. The engineers I spoke with were constantly monitoring their laptops as we talked, and rushing from meeting room to meeting room.

Nathan, the head of the product engineering team, said the focus remains on “the promise of the magic box, but I still think there’s too much complexity…I’m very optimistic that we can solve it, with the model and in a truly AI-native way.”

General Motors electric vehicles, including ones built in partnership with Honda, are now facing increased scrutiny from the top U.S. auto safety regulator after hundreds of incidents, more than 20 crashes or fires, and at least six injuries.

The brake problems also extend to some non-EV models, including the Chevy Colorado, GMC Canyon, and the Buick Enclave and Envision. More than 1 million vehicles may be affected.

The National Highway Traffic Safety Administration (NHTSA) started its investigation in April 2024 after reports of trouble from owners of 2023 model year Cadillac Lyriq vehicles. The agency’s Office of Defects Investigation (ODI) said Monday that it was upgrading this probe to what’s known as an “engineering analysis.” That’s the highest level of investigation that ODI performs, and is often a step the office takes before telling a company to issue a recall.

The initial complaints ODI received two years ago typically involved owners describing receiving a “Brake System Failure” message when starting up the vehicle, or after coming to a complete stop. GM performed “several internal investigations” into the issue, according to ODI, and determined that the problem was linked to fractures in the spindle of its “eBoost” brake-by-wire system.

But ODI said on Monday that it kept receiving reports of a loss of braking assistance that were “inconsistent with GM’s description of a spindle failure.” The additional reports described an “immediate loss of brake assist” while a customer was in the process of braking to slow their car down, which the safety regulator said “could result in extended braking distance, which increases the risk of a crash or injury.” ODI said it needs to do a further analysis of the potential for failures in the eBoost system.

“The safety of our customers is the highest priority for the entire GM team,” the automaker said in a statement. “We take these reports seriously and have conducted a thorough investigation, determining the alleged condition does not present an unreasonable risk to motor vehicle safety. We will continue to cooperate with NHTSA as its investigation proceeds.”

The eBoost system was introduced in 2019 and gradually rolled out to more models over the years. The system ditches a traditional mechanical link between the brake pedal and the braking system, opting for an electronic one instead. This allows GM to change the brake “feel” in different driving modes. The automaker put eBoost on its most popular EVs, like the Blazer EV, Equinox EV, Cadillac Lyriq, and the Honda Prologue and Acura ZDX, which it made with Honda in a joint venture. The Cruise Origin — the purpose-built electric autonomous vehicle with no steering wheel or pedals, which GM abandoned in 2024 — also used eBoost.

In one crash reported to NHTSA, the driver of a 2025 Lyriq said they lost their brakes while trying to pull into a parking space in front of the store. The vehicle drove over the curb and crashed through the store front, coming to rest “mid-way in the store, amidst the furniture and store structure,” according to the driver.

In another, the driver of a 2024 Blazer EV said they had to “deliberately steer the vehicle into a concrete curb” to slow it down and avoid a “catastrophic intersection collision.”

This story has been updated with a statement from GM.

We’re getting close to TechCrunch Disrupt 2026! In less than two months, more than 10,000 innovators, founders, investors, and industry leaders will gather in San Francisco. You can meet up with that community the week of TechCrunch Disrupt, which runs from October 13-15.

With our Disrupt Side Events program, you can bring people together anytime from October 10-16 for a custom event that gets promoted to Disrupt’s attendees. But time is running out — our deadline to apply for Side Events is September 4, so act now!

What are the benefits of a Side Event?

Hosting a Side Event puts you in control of how your community meets up, placing your brand front and center with the people who matter most. Whether you’re a VC looking to spark conversations, a company ready to showcase breakthrough innovation, or a community builder with something to say, this is your moment to:

Engage directly with thousands of the sharpest minds in tech.

Build authentic relationships with potential partners, investors, and collaborators.

Own your narrative and showcase your expertise on your terms.

Generate buzz that extends far beyond the event itself.

Approved Side Events are all promoted to Disrupt attendees and the rest of our audiences through the official Side Events calendar and marketing channels.

What Side Events can you host?

Anything that aligns with your vision is likely to be approved. Everything from industry roundtables and VC office hours to deep-dive workshops to intimate dinners and casual networking happy hours have succeeded at past Disrupt events.

The format is yours to design — we just ask that you bring the energy and make sure it’s open to both Disrupt attendees and the broader public. We recommend hosting events that take place in the evenings during Disrupt from October 13-15, as there’s less competition and more post-conference excitement to fuel meaningful conversations with undivided audiences.

How can I apply for a Side Event?

You can look at the application form here, but it’s helpful to have a few things locked in before starting the application process:

Have a venue selected in the Bay Area — this is all about in-person experiences.

Have a title, description, and venue prepared for the event.

Double-check the Side Events Host Guide and the Side Event Host T&Cs.

If you have additional questions, we’re here to help! Just email our Side Events team and we’ll follow up to help.

We prioritize events that provide meaningful value to founders, investors, and the startup community, but aside from that, you know your community and ideal audience the best! There’s zero cost to apply and participate, so let’s hear your ideas.

Apply by September 4 to be a part of Disrupt

Apply now and tell us your vision, goals, and what you need logistically in your application, and we’ll handle the rest to give your team the support and exposure you need to make the most of the community and opportunity Disrupt brings to San Francisco and the Bay Area.

美國司法部宣布，TikTok、母公司ByteDance及關係企業同意支付4億美元，和解2024年提出的兒童隱私訴訟。案件涉及《兒童線上隱私保護法》（COPPA），美國政府指控TikTok在未取得家長同意下蒐集及保存13歲以下兒童的個人資料。司法部表示，本案追償金額位居COPPA案件歷來前列。

依和解內容，TikTok將先支付3億美元，另1億美元則待法院撤銷2019年TikTok前身Musical.ly案件的既有法院命令後支付。

美國司法部與聯邦貿易委員會（FTC）2024年提告時指控，TikTok明知部分使用者未滿13歲，仍允許兒童建立一般TikTok帳號，並在未通知家長及取得同意的情況下蒐集與保存個資。即使是為13歲以下使用者設計的兒童模式，TikTok也被指在問題回報功能中蒐集兒童的電子郵件地址。

訴訟同時涉及兒童資料刪除機制，美國政府指控，部分家長要求刪除兒童帳號及相關個資時，TikTok未依要求完成處理，用來辨識及移除兒童帳號的內部流程也有不足。

COPPA要求線上服務若明知使用者未滿13歲，在蒐集、使用或揭露兒童個資前，須先通知家長並取得同意，也必須在家長要求時刪除已蒐集的兒童個資。TikTok前身Musical.ly早在2019年就曾因兒童隱私問題，以570萬美元和解FTC指控，並受到法院命令約束，後續必須遵守COPPA。

司法部表示，自2024年提告以來，TikTok在所有權、管理、法令遵循及隱私措施方面已有重大變化，並加強年輕使用者保護、年齡控管及家長監督機制，相關變化也納入此次和解考量。司法部同時強調，雙方以和解方式結束訴訟，法院並未就政府提出的指控作成違法或責任認定。

行動資安業者Zimperium發現新版Android銀行木馬ToxicPanda 2.0，攻擊者可透過167種遠端指令控制受感染手機，並針對分布在16國的349款銀行、金融、電子錢包及加密貨幣應用，顯示仿冒登入或交易畫面，竊取帳密與個人識別碼（PIN）。惡意程式還能偽造Android鎖定畫面，取得手機解鎖PIN碼、圖形或密碼。

ToxicPanda於2024年首次被揭露，當時已控制超過1,500臺Android裝置，受害情況主要集中在義大利、葡萄牙、西班牙及拉丁美洲。早期版本部分指令尚未具備實際功能，而新版已補上多項遠端控制與資料竊取能力。

感染流程由負責安裝的第一階段程式開始，惡意程式顯示網頁形式的假安裝畫面，誘使使用者授予VPN權限，藉此阻斷Google Play及Google Play Services的網路通訊，再從安裝檔內解開並安裝真正的惡意程式。接著，ToxicPanda會要求使用者開啟Android無障礙服務，藉此讀取畫面內容並代替使用者點選按鈕。

完成安裝後，ToxicPanda 2.0會盤點手機內的應用程式，將應用程式的套件名稱與圖示傳送至命令與控制伺服器。當受害者開啟目標金融應用，惡意程式就會下載外觀相似的假網頁，覆蓋原本的登入或交易畫面，攔截使用者輸入的帳密與PIN碼。

除了針對349款金融應用準備仿冒畫面，ToxicPanda 2.0另有一套鎖定超過140款銀行及加密貨幣應用的PIN碼竊取機制。惡意程式可在畫面上放置看不見的透明介面，攔截使用者的觸控操作，並可依照攻擊者下達的指令更換監控目標。

ToxicPanda 2.0也會利用無障礙服務，自動連點系統設定中的版本號碼開啟開發人員選項，再啟用Android偵錯橋接器（ADB）的無線偵錯功能，還會讀取手機顯示的6位數配對碼並自行完成連線，之後便能透過ADB取得shell使用者權限，以較高權限執行命令、替自己授予權限及調整系統設定。

其他功能包括顯示假系統更新畫面遮住背景操作、自動調整不同手機品牌的權限與電池最佳化設定，以及維持惡意程式持續執行，研究團隊也發現部分樣本透過AWS託管的儲存空間散布。

資安業者LevelBlue旗下SpiderLabs釋出開源攻擊測試框架RAVEN，整合針對Elasticsearch與Kibana的偵察與漏洞利用功能，並可進一步匯出Elasticsearch資料、建立持久化存取途徑。在取得Elasticsearch存取權後，RAVEN可建立不受原帳號密碼變更影響的API金鑰，甚至利用Elasticsearch內建排程機制，在額外帳號或API金鑰遭移除後重新建立。相關操作皆在Docker實驗環境進行。

RAVEN可分批讀取Elasticsearch索引內的文件，將指定索引或所有非系統索引匯出成檔案。大量資料下載中斷時，也能記錄進度並從上一個完成的批次繼續，不必重新取得全部內容。

除了直接讀取資料，RAVEN也能利用Elasticsearch原有的快照功能，在目標伺服器建立快照儲存庫，再把指定索引製作成快照。快照建立主要在Elasticsearch伺服器端完成，相較持續從外部大量查詢並傳輸資料，可減少長時間的網路傳輸活動。

RAVEN取得具有足夠權限的帳號後，還能建立新的Elasticsearch API金鑰。API金鑰會繼承建立者擁有的權限，但不會因原帳號更換密碼而自動失效，因此即使事件處理人員更換使用者密碼，既有金鑰仍可能繼續存取系統，必須另外找出並撤銷。

持久化功能則包含建立額外高權限帳號、長期有效的API金鑰，以及利用Elasticsearch內建的Watcher排程功能定期檢查兩者是否仍存在。測試範例設定為每小時檢查一次，如果帳號或API金鑰遭刪除，Watcher就會重新建立，因此只更換密碼、刪除額外帳號或撤銷金鑰，仍不足以完全移除相關存取途徑，還必須找出並移除對應的Watcher排程。

由於要避免滲透測試結束後留下系統變更，RAVEN會記錄建立API金鑰、帳號與快照儲存庫等操作，並保存對應的復原資訊，供測試人員後續清除。RAVEN目前已透過GitHub與Python套件庫PyPI提供。

中國迷你電腦品牌GEEKOM證實，舊版官方支援頁曾提供一個遭多個安全掃描服務判定為惡意程式的網路驅動程式安裝檔，其中部分掃描結果將其辨識為Asruex木馬。該套件適用於A7、A8、AE7、AE8、AX7 Pro及AX8 Pro等6款AMD迷你電腦。原廠已下架相關檔案，並強調問題只涉及使用者自行下載的驅動程式套件，出廠預載的Windows系統不受影響。

事件最初由Reddit使用者揭露，該使用者更換固態硬碟並重新安裝Windows後，從GEEKOM網站下載約1.1GB的整合驅動程式套件，解壓縮時收到Microsoft Defender警告。外媒VideoCardz隨後從相同官方網址取得檔案，利用多個掃描與分析服務複核，也得到惡意程式判定。

問題檔案位於整合套件的網路驅動程式資料夾，部分掃描結果將其辨識為Asruex。Asruex是已知的Windows木馬家族，具備後門功能，可讓攻擊者對受感染電腦執行遠端操作。不過，目前沒有完整技術分析，無法確認外媒提到的鍵盤側錄、密碼竊取等功能是否曾在GEEKOM迷你電腦實際運作。

GEEKOM表示，存放檔案的舊支援頁已由新版系統取代，無法從官網正常導覽進入，但頁面與下載網址仍被搜尋引擎收錄。公司檢查目前支援頁後未發現相同異常，已移除相關問題頁面與檔案，並將加強檔案管理及審查流程。

VideoCardz指出，相同檔案雜湊值的分析紀錄可追溯至2024年12月。雜湊值可視為辨識檔案內容的數位指紋，不過，GEEKOM目前仍未說明問題檔案如何進入官方驅動套件，也未公布檔案存放時間、下載次數及實際受影響使用者規模。

GEEKOM官方建議，曾下載問題網路驅動程式套件的使用者不要執行安裝檔，並刪除裝置上留存的檔案，同時以Microsoft Defender或其他可信任的防毒軟體完整掃描系統，再透過Windows Update、Realtek官網或GEEKOM新版支援頁重新安裝網路驅動程式，要是仍無法排除疑慮，可使用微軟官方提供的Windows安裝映像檔全新安裝系統。