Anthropic is asking investors to submit allocations for the AI company’s latest fundraise within the next 48 hours, according to sources familiar with the matter. The round, which TechCrunch reported is expected to be roughly $50 billion, is estimated to close within two weeks, the sources said.

As we previously reported, Anthropic is targeting a valuation of about $900 billion. However, given the soaring demand from investors seeking a stake in the company, the final valuation may well exceed that figure, our sources said.

Anthropic declined to comment.

Despite the intense demand, some early backers — particularly those who invested in 2024 or earlier — are skipping this round. Instead, these investors are waiting to potentially cash out during Anthropic’s anticipated IPO later this year.

The company is raising what is likely to be its last private round before going public to fund its massive computing needs.

Anthropic announced this month that its annual revenue run rate has surpassed $30 billion. But as we previously reported, the company’s run rate is currently closer to $40 billion, according to sources with knowledge of the company’s financials.

Anthropic raised its last round in February at a $380 billion valuation. At $900 billion, the company would not only more than double its valuation but would also surpass its chief rival, OpenAI, which closed a record-breaking $122 billion round at an $852 billion post-money valuation earlier this year.

Techcrunch event Meet your next investor or portfolio startup at Disrupt

Your next round. Your next hire. Your next breakout opportunity. Find it at TechCrunch Disrupt 2026, where 10,000+ founders, investors, and tech leaders gather for three days of 250+ tactical sessions, powerful introductions, and market-defining innovation. Register now to save up to $410. Meet your next investor or portfolio startup at Disrupt

Your next round. Your next hire. Your next breakout opportunity. Find it at TechCrunch Disrupt 2026, where 10,000+ founders, investors, and tech leaders gather for three days of 250+ tactical sessions, powerful introductions, and market-defining innovation. Register now to save up to $410. San Francisco, CA | REGISTER NOW

Jagmeet Singh contributed reporting.

Nvidia has laid a new brick in its AI empire. NVentures, its corporate VC fund, has backed Legora, reportedly its first legal AI investment.

Leveraging AI to help lawyers streamline their work, the Swedish-born legal tech startup is competing with U.S. player Harvey.

Alongside Atlassian and other new financial investors, NVentures joined Legora’s cap table as part of a $50 million Series D extension that comes a month after the startup’s $550 million Series D.

In the interval, this Y Combinator alum crossed $100 million in annual recurring revenue (ARR) — a milestone that contributed to its new $5.6 billion post-money valuation.

This brings Legora’s valuation just a tad closer to Harvey’s, which reached $11 billion last month when Sequoia tripled down on its investment. Andreessen Horowitz, Coatue, Conviction Partners, Elad Gil, Matt Miller’s Evantic, and Kleiner Perkins also participated in that round.

Legora, too, is backed by high-profile VCs, but it puts even more emphasis on the big names it secured as clients, such as Bird & Bird, Cleary Gottlieb, and Linklaters. According to the company, the platform it launched only 18 months ago is now used by more than 1,000 law firms and in-house legal teams across 50 markets.

Harvey has game in that area too. It claims 100,000 lawyers across 1,300 organizations as customers, ranging from global law firms like Hengeler Mueller and Latham & Watkins to corporate legal teams at companies like T-Mobile and Bridgewater.

Techcrunch event Meet your next investor or portfolio startup at Disrupt

Your next round. Your next hire. Your next breakout opportunity. Find it at TechCrunch Disrupt 2026, where 10,000+ founders, investors, and tech leaders gather for three days of 250+ tactical sessions, powerful introductions, and market-defining innovation. Register now to save up to $410. Meet your next investor or portfolio startup at Disrupt

Your next round. Your next hire. Your next breakout opportunity. Find it at TechCrunch Disrupt 2026, where 10,000+ founders, investors, and tech leaders gather for three days of 250+ tactical sessions, powerful introductions, and market-defining innovation. Register now to save up to $410. San Francisco, CA | REGISTER NOW

With global leadership as the end goal, the Harvey v. Legora rivalry is one they intend to play on each other’s home turf. Legora has opened multiple offices around the world with the U.S. a key focus for its expansion. Conversely, Harvey is pushing into Europe.

With plenty of capital to spend on both sides, that battle has moved to mindshare. Not long after Winston Weinberg’s company Harvey signed a brand partnership with actor Gabriel Macht, who plays a high-powered lawyer in the TV series “Suits,” Legora launched an advertising campaign featuring movie star Jude Law under the slogan “Law just got more attractive.”

Both companies may be right to bet heavily on marketing. Rivalry aside, they are built on top of large language models made by AI giants that could well become their competitors. When Anthropic launched a legal plug-in for Claude not long ago, several publicly listed legal software companies saw their stocks drop.

Legora CEO Max Junestrand says he isn’t concerned.

“Foundation models are improving quickly, but the real value is in how they’re applied,” he wrote in a statement. It also shows how the startup instills FOMO among its target users, stating that “the legal teams that embed AI effectively today will shape how the industry evolves.”

NVentures’ investment is also a signal that Legora might have enough of a moat to protect them from the model makers, and its bigger rival.

However, Nvidia is also known for hedging its bets — after all, it invested in both Anthropic and OpenAI before deciding it has probably had enough.

Rivian has reworked its loan deal with the Department of Energy and now expects to borrow $4.5 billion to build its new factory in Georgia, down from the original amount of $6.6 billion allocated under the Biden administration.

The company also announced Thursday that it will draw on the loan sooner than planned, in early 2027, and expects to increase the total capacity of the Georgia plant from 200,000 to 300,000 vehicles in its initial phase of operation — another sign that the company has high hopes for its upcoming R2 SUV.

The larger capacity — a 50% increase over its initial plans — will help lower its per unit costs, while also providing significant room for future expansion of capacity in later phases, the company said Thursday.

Some of the factory’s capacity will be used to produce R2 robotaxis for Uber. Under a deal struck earlier this year, Uber is making an initial $300 million investment in Rivian and is expected to purchase 10,000 fully autonomous R2 robotaxis ahead of a planned rollout in San Francisco and Miami in 2028. That initial $300 million payment is expected to close in the second quarter, and another $250 million investment is planned for later this year, according to Rivian.

The ride-hailing company has the option to buy up to 40,000 more autonomous R2 SUVs from Rivian starting in 2030. Uber has said it will invest up to $1.25 billion in Rivian through 2031 if the automaker meets a series of milestones.

Rivian broke ground on the Georgia factory late last year and is in the beginning stages of doing so-called vertical construction at the site located outside Atlanta. The company expects to start making vehicles by the end of 2028. Until then, Rivian will build R2 SUVs at its current factory in Normal, Illinois.

The company recently started production of the R2 despite the plant suffering damage from a tornado, and Rivian said Thursday it has made initial deliveries to employees. Deliveries to customers are expected to start “in the coming weeks,” according to Rivian.

Techcrunch event Meet your next investor or portfolio startup at Disrupt

Your next round. Your next hire. Your next breakout opportunity. Find it at TechCrunch Disrupt 2026, where 10,000+ founders, investors, and tech leaders gather for three days of 250+ tactical sessions, powerful introductions, and market-defining innovation. Register now to save up to $410. Meet your next investor or portfolio startup at Disrupt

Your next round. Your next hire. Your next breakout opportunity. Find it at TechCrunch Disrupt 2026, where 10,000+ founders, investors, and tech leaders gather for three days of 250+ tactical sessions, powerful introductions, and market-defining innovation. Register now to save up to $410. San Francisco, CA | REGISTER NOW

The modifications to the DOE loan come as Rivian revealed financial results for the first quarter of 2026 on Thursday. The company generated $1.38 billion in revenue, with $908 million coming from vehicle sales and $473 million from software and services. Rivian’s automotive revenue declined about 2% from the same year-ago period, due in part to a drop in regulatory credits.

The company lost $416 million in the quarter, down from a $541 million loss in the same period last year. That net loss shrank thanks, in part, to a $506 million gain in other income related to the Series A capital raise and related deconsolidation of CEO RJ Scaringe’s new startup Mind Robotics, according to the company.

Rivian saw its operating expenses and R&D costs grow year-over-year. Rivian’s R&D budget expanded 20% to $458 million as it increased spending on R2 pre-production costs as well as software and cloud services related to the development of autonomous vehicle technology.

The combination of these rising costs, plus a small uptick in capital spending, was a drag on Rivian’s free cash flow, which is in negative territory. The company reported a negative free cash flow of $1 billion, nearly double from a year ago.

In Brief

VC firm 137 Ventures announced Thursday that it has raised more than $700 million across two new growth-stage funds to back startups “with the potential for significant market impact.”

In the past year, the firm, founded by former Founders Fund investor Justin Fishner-Wolfson, said it has deployed more than a billion dollars into companies building in defense, AI, and industrial systems. Its portfolio includes AI agent company Cognition, AI-powered manufacturing company Hadrian Automation, and defense company Anduril.

It is also, notably, an investor in Elon Musk’s SpaceX, which is expected to have a historic IPO this year that could value the company at more than a trillion dollars. Bloomberg reported that 137 Ventures first backed SpaceX in 2010 and has since cut the company around two dozen checks.

Practice by Numbers, the developer of a patient management software used in thousands of dentists’ offices, has fixed a security flaw that exposed the private health records of patients on a portal that comes bundled with the software, TechCrunch has learned.

One patient, Joseph R. Cox, reported the bug to TechCrunch after he encountered the issue while looking at his own dental records on the portal, which was offered by his dentist’s office.

This patient portal is part of a dental office management software made by Practice by Numbers, which claims its products are used in over 5,000 dental practices across the United States.

Cox said the bug allowed any user of the portal, which houses patients’ medical documents and health records, to access documents belonging to other patients. He said he was able to access other patients’ documents from his account, including their personal information, medical histories, photo identification, and other files. The bug also meant that Cox’s records were just as exposed to other patients.

Cox said he attempted to alert the company about the issue via email, but did not hear back. He then notified TechCrunch as a last resort to ask the company to patch the bug.

The bug was remarkably easy to exploit by anyone with a login to the Practice by Numbers’ patient portal. Cox said changing the document number in the web address while loading one of his documents in the portal allowed users to access other patients’ files.

Worse, Cox said the document numbers in the web address appear to be sequentially incremental, so it could be possible to easily guess the document numbers of other people’s medical files.

Cox told TechCrunch that he faced difficulties in alerting Practice by Numbers to the issue, as the company offered no discernible avenue to report security problems. The company’s email address on its website was broken, with emails returned as undeliverable. Instead, Cox sent a message to one of the company’s founders on LinkedIn, but heard nothing back after sending a subsequent email.

The issue, now fixed, highlights a recent trend in which regular consumers are finding security flaws in companies’ products or websites, but have no clear way to report the issue to the developers.

Earlier in April, fashion retailer Express fixed a website bug that allowed anyone to access the order details and personal information of other customers, after a user identified the bug, but found no way to alert the company. A similar incident involved Home Depot in December: A security researcher tried to privately alert the company about a security lapse that was exposing access to its internal systems for almost a year, but their reports were ignored until TechCrunch contacted the company.

Given the security flaw was actively putting patients’ data at risk, TechCrunch alerted Practice by Numbers to the issue on April 13. The company took down its patient portal to fix the bug, and brought it back online on April 17.

Practice by Numbers’ co-founder and chief technology officer, Chris Lau, told TechCrunch that the company had fixed the vulnerability, and it was notifying fewer than 10 patients that their information was exposed due to the bug, citing its server logs.

The company said it was working with the affected dental practice to notify the affected patients. Lau said that the company had not identified evidence of previous activity related to the bug, suggesting Cox was likely the first to find it.

Cox confirmed that the bug appears to have been fixed.

When asked by TechCrunch, neither Lau nor Practice by Number’s co-founder and president, Rohit Garg, would say if the company’s patient portal had undergone a security audit before it was launched. Companies commonly undergo security audits to ensure their products meet cybersecurity standards, and are free from common security flaws before customers begin using them.

While no software is ever completely bug-free, companies that handle sensitive information, like healthcare data, typically seek third-party reviews of their code to weed out any major security flaws.

When asked if Practice by Numbers plans to update its website to allow security researchers to notify the company of security flaws, such as through a vulnerability disclosure program, Garg said the company plans to update its website to let people report security issues. The company did not offer a timeline.

Security researchers are sounding the alarm on a newly discovered vulnerability in the widely used web server management software cPanel and WebHost Manager (WHM).

The bug allows hackers to hijack and take full control of the servers running the affected software, which is thought to be used by tens of millions of website owners around the world.

Many commercial web hosting companies have patched their customers’ systems already. But the cPanel maker urged customers to ensure that their systems are patched as the bug affects all supported versions of the software.

cPanel and WHM are two software suites used for managing web servers that host websites, manage emails, and handle important configurations and databases needed to maintain an internet domain. The two suites have deep-access to the servers that they manage, allowing a malicious hacker potentially unrestricted access to data managed by the affected software.

The bug, officially tracked as CVE-2026-41940, allows malicious hackers to remotely bypass its login screen to gain full access to the software’s administration panel.

Given the ubiquity of the cPanel and WHM software across the web hosting industry, hackers could compromise potentially large numbers of websites that haven’t patched the bug.

Canada’s national cybersecurity agency said in an advisory that the bug could be exploited to compromise websites on shared hosting servers, such as large web hosting companies.

The agency said that “exploitation is highly probable” and that immediate action from cPanel customers, or their web hosts, is necessary to prevent malicious access.

Web hosting giant Namecheap, which uses cPanel to allow its customers to manage their web servers, said the company blocked access to customers’ cPanel panels after learning of the flaw to prevent exploitation, and to give it time to patch its customers’ systems.

HostGator also said it patched its systems and is considering the bug a “critical authentication-bypass exploit.”

One web hosting company says it found evidence that hackers have been abusing the vulnerability for months before the attempts were discovered.

KnownHost CEO Daniel Pearson said in a post on Reddit that his company has seen attempts to exploit the vulnerability as far back as February 23. The company said it also briefly began blocking access to customer systems before applying patches.

According to Pearson, around 30 servers at KnownHost showed signs of unauthorized attempted access out of thousands of computers on its network. Pearson likened the efforts to attempts, and has not seen signs of active compromise. cPanel also said it rolled out a security fix for WP Squared, a similar tool for managing WordPress websites.

In Brief

After Sam Altman trash-talked Anthropic for gatekeeping its cybersecurity tool Mythos by only releasing it to select users, he confirmed that OpenAI would be doing the same with its competing tool, Cyber.

Altman said in a post on X on Thursday that OpenAI will begin rolling out GPT-5.5 Cyber “to critical cyber defenders” in the next few days. OpenAI has an application on its website where people submit information about their credentials and planned use in order to gain access.

Cyber can perform such tasks as penetration testing, vulnerability identification (and exploitation), and malware reverse engineering, the application implies. It’s intended to be a toolkit to help a company find security holes and test defenses. The fear is that the kit could be misused by the bad guys.

When Anthropic similarly restricted access to Mythos, Altman called the tactic fear-based marketing. Some critics also thought so, saying Anthropic’s rhetoric was overblown. Ironically, an unauthorized group reportedly managed to gain access to Mythos anyway.

OpenAI says it’s working to make Cyber more widely available by consulting with the U.S. government and identifying more users with legit cybersecurity credentials.

Faraday Future paid around $7.5 million to a company controlled by its founder Jia Yueting in 2025, according to a new regulatory filing.

The long-struggling electric vehicle startup made the payments in a year when it delivered only four vehicles and lost nearly $400 million. The company has pivoted to selling cheaper vans and robots imported from China.

The payments happened while Faraday Future was still under investigation by the Securities and Exchange Commission (SEC), which was probing what are known as “related party transactions” between the company and entities related to or controlled by Jia, Faraday’s own filings have shown. The SEC was also investigating whether Faraday Future properly represented the level of control Jia had over the company when it went public in 2021, and whether it lied about early sales of its EVs in 2023.

The SEC dropped its four-year investigation in March, as TechCrunch first reported, despite having sent notices to Faraday Future, Jia, and other executives last year stating that investigators were recommending an enforcement action. The closure of the investigation comes amid a historic drop in white-collar crime enforcement during the second Trump administration.

The new transactions were revealed in Faraday Future’s annual proxy filing published on Thursday. It shows Faraday Future paid a mix of monthly $100,000 “consulting” fees, a $2 million “bonus payment,” and $1.7 million to repay loans from the company, which is called FF Global Partners LLC. The company did not explain the remaining $2.6 million in the filing.

Faraday Future did not respond to a request for comment.

Faraday Future describes FF Global as an “affiliate” of Jia in the proxy filing, and in previous filings has said he exerts “significant influence” over the LLC. FF Global has five “voting managers,” one of whom is Jia, while the others include business associates and a family member — his nephew Jerry Wang.

Techcrunch event Meet your next investor or portfolio startup at Disrupt

Your next round. Your next hire. Your next breakout opportunity. Find it at TechCrunch Disrupt 2026, where 10,000+ founders, investors, and tech leaders gather for three days of 250+ tactical sessions, powerful introductions, and market-defining innovation. Register now to save up to $410. Meet your next investor or portfolio startup at Disrupt

Your next round. Your next hire. Your next breakout opportunity. Find it at TechCrunch Disrupt 2026, where 10,000+ founders, investors, and tech leaders gather for three days of 250+ tactical sessions, powerful introductions, and market-defining innovation. Register now to save up to $410. San Francisco, CA | REGISTER NOW

Wang, who is a president at Faraday Future, draws a six-figure salary from FF Global, according to the filing. So does his wife, who is the head of FF Global’s legal department. FF Global also has a similar “consulting agreement” with a crypto holding company run by Wang (and advised by Jia) called AIXC. (Wang’s wife’s law firm also consults for AIXC.)

FF Global is also a major shareholder of Faraday Future and — with Jia — controls almost every aspect of the EV company, to the point that Faraday labels this as a risk factor in its most recent annual filing.

“Jia and FF Global, over which Mr. Jia exercises significant influence, have control over our management, business and operations, and may use this control in ways that are not aligned with our business or financial objectives or strategies or that are otherwise inconsistent with our interests,” the company wrote earlier this year.

FF Global also helped bring Jia back to power after the company went public in 2021. Shortly after Faraday Future merged with a special purpose acquisition company, the new public company board of directors opened an investigation into Jia’s movement of money in and out of the company, and into the disclosures made during the merger process.

In early 2022, the board sidelined Jia, who has been blacklisted by China for financial fraud, after finding Faraday Future had misrepresented the level of control he had over the company. They then referred their findings to the SEC, which opened its investigation shortly after.

FF Global, meanwhile, spent all of 2022 agitating to replace certain board members with ones friendly to Jia. This campaign became so intense that multiple board members received death threats. Those board members ultimately resigned in part because they feared for their lives. Jia was re-installed as co-CEO last year, and is now Faraday Future’s sole CEO.

FF Global is not the only Jia-related company that Faraday Future has paid, or plans to pay, money to. The company stated in its proxy filing that it paid $700,000 to a loan company associated with him last year. It also owes $8.5 million to Leshi Information Technology Co. Ltd., one of the companies related to his failed Chinese tech conglomerate LeEco, for “advertising services.”

OpenAI is getting serious about account security.

The company on Thursday launched Advanced Account Security (AAS), a set of opt-in protections for ChatGPT users designed for high-value individuals — but available to anyone who wants them.

As part of that new program, digital security provider Yubico announced it has partnered with OpenAI to link two new security key products to ChatGPT accounts. The company said the partnership was designed to protect users from the threat of phishing, which is considered to be a growing threat for chatbot users.

The two companies are releasing a pair of “co-branded” YubiKeys — dubbed the YubiKey C NFC and the YubiKey C Nano.

OpenAI has suggested that AAS is a good fit for political dissidents, journalists, researchers, and elected officials — people who engage in politically charged and risky work. One would assume that it might make sense for enterprise users, whose corporate secrets are squirreled away in ChatGPT sessions.

“Ultimately, our intent is to drastically reduce the threat of unauthorized access to sensitive data in OpenAI accounts worldwide,” Yubico CEO Jerrod Chong said in press release announcing the deal.

Security keys are small pieces of hardware that can be tied to digital accounts and enacted through a computer’s USB ports. A unique cryptographic identifier lives on the key, which allows only the person in possession of it to log into a connected account.

Techcrunch event Meet your next investor or portfolio startup at Disrupt

Your next round. Your next hire. Your next breakout opportunity. Find it at TechCrunch Disrupt 2026, where 10,000+ founders, investors, and tech leaders gather for three days of 250+ tactical sessions, powerful introductions, and market-defining innovation. Register now to save up to $410. Meet your next investor or portfolio startup at Disrupt

Your next round. Your next hire. Your next breakout opportunity. Find it at TechCrunch Disrupt 2026, where 10,000+ founders, investors, and tech leaders gather for three days of 250+ tactical sessions, powerful introductions, and market-defining innovation. Register now to save up to $410. San Francisco, CA | REGISTER NOW

If the threat of phished ChatGPT accounts may seem somewhat abstract, there is a growing body of literature showing that bad actors are increasingly targeting chatbot users. Cybercriminals are always on the lookout for extortion-worthy information and, given the intimate nature of most chatbot conversations, there is plenty of fodder when it comes to both enterprise and personal-level users.

Digital security is also becoming a bigger focus of the AI industry. Several weeks ago, Anthropic announced a new cybersecurity model called Mythos. Perhaps seeking to steal some of its competitor’s thunder, OpenAI has also made a number of announcements related to digital security. Thursday’s news of the Yubico partnership followed OpenAI’s announcement that it’s launching a new framework for digital defense.

Of course, a security-key-enabled account does offer stronger protection, but it comes with a tradeoff: If the key is lost, OpenAI won’t be able to help recover access. In practice, that means conversations could be lost for good.

Apple’s iPhone sales and Services revenue were the stars of the show in the tech giant’s most recent quarter, but the Mac quietly outperformed — helped by growing demand for AI workloads.

Wall Street investors had expected to see Mac revenue in the low $8 billion range, but Apple reported $8.4 billion in the second quarter ended March 28 — a notable beat for a non-core segment of the tech giant’s business. In addition, investors ahead of earnings believed that Mac sales would be essentially flat year-over-year. Instead, Mac sales were up 6% on an annual basis, the company told investors. The company’s total revenue was $111.2 billion, a 17% increase from the same period last year.

Apple chalked up some of the Mac growth to recent product launches, including the well-received MacBook Neo. However, those fun, colorful computers were only on sale for a few weeks after the March 4 preorders began. Realistically, most units shipped mid- to late March, and some demand may have been pushed into April as certain models sold out.

Apple CEO Tim Cook told analysts on the company’s Q2 earnings call on Thursday that customer demand for the Neo was “off the charts” and higher than Apple had expected. He also noted that Apple set a record in the quarter for customers new to the Mac, partly due to the Neo.

Cook attributed the Mac sales growth to the use of the platform for running local AI models, like OpenClaw — something that took Apple somewhat by surprise as Mac mini and Mac Studio devices sold out in recent weeks.

“Both of these are amazing platforms for AI and agentic tools, and the customer recognition of that is happening faster than what we had predicted, and so we saw higher than expected demand,” Cook said of these Mac sales. He also noted that the Mac mini was the top-selling desktop in China — a market that’s been in an OpenClaw frenzy as of late.

Still, Mac revenue was flat on a quarter-over-quarter basis, suggesting this new demand has yet to scale. Cook said it may take Apple “several months” to reach supply-demand balance on the Mac mini and Studio models.

Techcrunch event Meet your next investor or portfolio startup at Disrupt

Your next round. Your next hire. Your next breakout opportunity. Find it at TechCrunch Disrupt 2026, where 10,000+ founders, investors, and tech leaders gather for three days of 250+ tactical sessions, powerful introductions, and market-defining innovation. Register now to save up to $410. Meet your next investor or portfolio startup at Disrupt

Your next round. Your next hire. Your next breakout opportunity. Find it at TechCrunch Disrupt 2026, where 10,000+ founders, investors, and tech leaders gather for three days of 250+ tactical sessions, powerful introductions, and market-defining innovation. Register now to save up to $410. San Francisco, CA | REGISTER NOW

“We’re not at the point where we’re saying this [constraint] is going to end anytime soon. And it’s not because of a problem, per se, other than we just under-called the demand,” Cook explained.

Enterprise demand for the Mac was also at play. Apple pointed to a couple of larger companies, including Perplexity, that had turned to Mac as their preferred platform for building enterprise-grade AI assistants.

He also said Apple was “supply constrained on the MacBook Neo,” and has even seen school systems, like Kansas City Public Schools, dropping Chromebooks for the Neo.

Founders building in the healthcare space can’t just build fast and break things. Timelines stretch longer, stakes are higher, and success depends on navigating systems that reward rigor over speed.

That’s exactly the reality Robhy Bustami, co-founder and CEO of BioticsAI, has been building in. His company is developing an AI copilot for ultrasound that helps detect fetal abnormalities, an area where misdiagnosis rates remain surprisingly high. Bustami joined Isabelle Johannessen on Build Mode to discuss how the company has navigated a highly regulated space and kept the team motivated while cutting through all the red tape.

BioticsAI started scrappy. The team built an early, functioning version of the product for under $100,000, an almost unheard-of milestone in the medical device world. That prototype helped them win TechCrunch Startup Battlefield in 2023, bringing early visibility and credibility. In January, they gained FDA approval, which means they can begin launching in hospitals and growing the business at a new rate.

From day one, the team approached product development with FDA approval in mind. Instead of building first and figuring out regulation later, they integrated clinical validation, regulatory strategy, and product development into a single process. That meant working closely with clinicians, collecting large-scale datasets, and running structured clinical studies before ever reaching the submission stage.

The FDA process itself is often viewed as a black box, but Bustami emphasizes that founders don’t have to navigate it blindly. Early engagement with regulators, through pre-submission meetings, helped the team align on study design and expectations. Still, risk never fully disappears. For many investors, the biggest question is simple: What if the FDA says no?

Internally, those long timelines create a different kind of challenge: keeping a team motivated when the biggest milestone is years away. At BioticsAI, that meant building a culture of alignment across engineers, clinicians, and researchers, ensuring everyone got to see the wins that were happening.

“Making sure everyone is completely aligned, even if it’s outside of their technical scope,” Bustami said, “constantly seeing wins on the R&D side,” from clinical studies to new healthcare partnerships.

Techcrunch event Meet your next investor or portfolio startup at Disrupt

Your next round. Your next hire. Your next breakout opportunity. Find it at TechCrunch Disrupt 2026, where 10,000+ founders, investors, and tech leaders gather for three days of 250+ tactical sessions, powerful introductions, and market-defining innovation. Register now to save up to $410. Meet your next investor or portfolio startup at Disrupt

Your next round. Your next hire. Your next breakout opportunity. Find it at TechCrunch Disrupt 2026, where 10,000+ founders, investors, and tech leaders gather for three days of 250+ tactical sessions, powerful introductions, and market-defining innovation. Register now to save up to $410. San Francisco, CA | REGISTER NOW

Now, with FDA clearance secured, BioticsAI is entering a new phase: deployment. The company is beginning to roll out its technology in hospitals, with plans to expand beyond obstetrics into broader areas of reproductive health.

Loading the player…

Building in healthcare is a long game. It requires patience, discipline, and a willingness to operate in uncertainty. For founders willing to take that path, the reward isn’t just a successful company — it’s the chance to build something that genuinely changes how care is delivered.

Subscribe to Build Mode on Apple Podcasts, Spotify, or wherever you like to listen. Watch the full videos on YouTube. Isabelle Johannessen is our host. Build Mode is produced and edited by Maggie Nye. Audience Development is led by Morgan Little. And a special thanks to the Foundry and Cheddar video teams.

Apply to Startup Battlefield: We are looking for early-stage companies that have an MVP. So nominate a founder (or yourself). Be sure to say you heard about Startup Battlefield from the Build Mode podcast. Apply here.

TechCrunch Disrupt 2026: We’re back for TechCrunch Disrupt on October 13 to 15 in San Francisco, where the Startup Battlefield 200 takes the stage. So if you want to cheer them on, or just network with thousands of founders, VCs, and tech enthusiasts, then grab your tickets.

Use code buildmode15 for 15% off any ticket type.

Google announced on Thursday that it will begin rolling out Gemini to cars with Google built-in, marking a significant upgrade from the current Google Assistant. The move signals Google’s push to bring more advanced, conversational AI into the driving experience.

The announcement follows closely behind news from General Motors, which revealed yesterday that Gemini is coming to approximately 4 million vehicles from model year 2022 and newer, spanning brands like Cadillac, Chevrolet, Buick, and GMC. However, today’s announcement didn’t name specific automakers, suggesting that Gemini won’t be limited to GM vehicles.

The rollout will begin in the U.S. with English-language support, with availability expanding over the coming months. Notably, Gemini isn’t just reserved for new vehicles, but also for compatible existing cars via software updates.

Cars with Google built-in first launched in 2020. Now, according to Google, the new Gemini update enables a more natural, conversational way for drivers to interact with their vehicles.

Soon, drivers will be able to speak more freely to complete tasks, explore ideas, or retrieve information. For instance, a driver could say they want to stop for lunch at a highly rated sit-down restaurant with outdoor seating along their route. Gemini can pull relevant suggestions using data from Google Maps, then handle follow-up questions like parking availability or menu options, including dietary preferences.

Gemini can also complete tasks like turning on the heat, provide directions, recommend music, retrieve vehicle information, summarize incoming messages, and help drivers respond hands-free.

Additionally, there’s Gemini Live, which is currently in beta and allows for more open-ended, real-time conversations. Drivers can activate it by tapping a button in the interface or by saying, “Hey Google, let’s talk,” enabling brainstorming, learning, or general discussion while on the road.

Techcrunch event Meet your next investor or portfolio startup at Disrupt

Your next round. Your next hire. Your next breakout opportunity. Find it at TechCrunch Disrupt 2026, where 10,000+ founders, investors, and tech leaders gather for three days of 250+ tactical sessions, powerful introductions, and market-defining innovation. Register now to save up to $410. Meet your next investor or portfolio startup at Disrupt

Your next round. Your next hire. Your next breakout opportunity. Find it at TechCrunch Disrupt 2026, where 10,000+ founders, investors, and tech leaders gather for three days of 250+ tactical sessions, powerful introductions, and market-defining innovation. Register now to save up to $410. San Francisco, CA | REGISTER NOW

Drivers who are signed into their Google accounts in compatible vehicles will be prompted with the option to upgrade. Once enabled, Gemini can be accessed through voice commands, the on-screen microphone, or steering wheel controls.

Google says it plans to expand Gemini support to additional languages and regions. Future updates are also expected to deepen integration with services like Gmail, Google Calendar, and Google Home.

TikTok is launching a new “Campus Hub,” the company told TechCrunch in an email on Thursday. The new hub features dedicated college group chats and personalized feeds designed to help students stay connected with their campus communities, even while they’re away for the summer.

The new hub builds on the app’s campus verification feature, which launched last August and lets users add their college campus to their TikTok profile and browse a list of students at their school in order to find and connect with others on their campus. The feature is available across more than 6,000 universities through TikTok’s partnership with UNiDAYS, a student verification platform that confirms student status.

Once users confirm their student status, they will get access to their new dedicated campus hub.

Students can create and join group chats with up to 300 classmates. These group chats are exclusive to verified students at the same school. TikTok says these chats can be used to stay in touch, plan reunions, or keep conversations going over the summer.

Image Credits:TikTok /

While these chats could be useful for the summer, TikTok may also be aiming to get students to use its app for everyday school communication, like class or club discussions that typically take place on other platforms such as Discord, Facebook Messenger, or Instagram.

As for the new college feeds, users will see a personalized feed that surfaces a mix of content posted by verified students and content related to their university. TikTok says these feeds will allow students to stay plugged into campus life, trends, and updates from wherever they may be.

The new features are reminiscent of Facebook’s early days, when the platform was built entirely around college campuses and required people to sign up with a valid “.edu” email address in order to connect with classmates and meet people at their university.

It’s worth noting that TikTok isn’t the only social media giant looking to make it easier for college students to connect with each other, as Meta-owned Instagram rolled out a feature last year that’s nearly identical to TikTok’s campus verification feature, as it also allows students in the U.S. to add their college or university to their profile banner and browse a list of students at their school.

Financial services platform Stripe is introducing a digital wallet specifically built for the AI era, where autonomous agents can perform tasks that include shopping, paying for reservations, buying tickets, and more.

At its annual conference this week, the company introduced Link, a wallet that allows you to connect various payment methods, track your spending, and view your recurring subscriptions. It also lets you integrate your AI agents so they can spend on your behalf, securely.

Link, which is available on the web, iOS, and Android, offers many things that you might expect from a digital wallet. You can connect payment methods, including cards, banks, crypto wallets, and buy now/pay later services, as well as store important details for online checkout like your billing and shipping information.

The wallet offers other handy features, too, like the ability to see what you’re spending, and track your recurring subscriptions — even updating the payment method the services have on file, as needed. It also offers 90 days of protection on eligible purchases from select merchants.

But what makes Link interesting is its ability to work with autonomous AI agents, like OpenClaw and others.

Image Credits:Stripe

The number of people experimenting with autonomous AI has been booming, to the point that Apple sold out of its base model Mac Minis, a popular platform for running these new, always-on AI agents. However, some people (rightly) take pause at the idea of giving an agent raw payment information, even if it’s providing convenience by automating various bookings.

Link is meant to offer a solution, as users can connect their AI agents and give them permission to pay, without exposing their payment credentials.

To work, users will first grant their agent access to the Link wallet via an OAuth (standard authentication) flow. The agent can then create a spend request, provide you with the context, and wait for your approval. Currently, it works with traditional payment methods, but Stripe says support for agentic tokens, stablecoins, and other types of payments is coming “soon.”

Image Credits:Stripe

On mobile and web, users will get a notification to approve the spend request, which requires they first review the transaction before the payment credential is shared with the AI agent. In the future, Stripe says it will expand its controls so users can set their own spending limits, or even choose when their agents can act without approval.

The wallet has been built on Stripe’s new Issuing for agents, which lets users issue virtual cards for agents to use to make purchases autonomously, with real-time authorization, spending controls, and full transaction visibility. Instead of giving an agent access to your payment credentials, users can either give agents programmatic access to Link, which provides a one-time-use card, or they can use a Shared Payment Token (SPT), which is backed by payment cards and banks.

Developers and businesses building agents or AI personal assistants can also use Link’s wallet instead of building their own wallet from scratch, Stripe notes.

Artificial intelligence continues to advance at a dizzying clip, forcing enterprises to develop and release new products quicker than ever or risk becoming irrelevant to a faster-moving competitor.

Salesforce believes it has found a strategy that allows it to keep up even if it isn’t clear where AI is headed next. The customer management software giant is crowdsourcing its AI roadmap in real time.

Salesforce is certainly not the only company to work intimately with its customers for feedback on its products. However, it’s notable considering the sheer size of the company, the pace of new product launches or fixes to existing ones, and the granular level of these relationships. These aren’t annual or even quarterly discussions. Salesforce is meeting with some customers as often as once a week.

“The 18,000 customers are a wellspring of information and a wealth of information that is really needed to get to customer success,” Jayesh Govindarajan, executive vice president at Salesforce AI, told TechCrunch in a recent interview. “The stack that we’ve built has resonated with these customers. Over time we can get context to be better, and as it gets better, and LLMs get better, agent systems do more and more fully autonomous behaviors. That’s a long-running innovation track and we’re going to invest in that.”

Salesforce was one of the first companies to launch AI agent management software in late 2024 before agentic AI started to dominate headlines the following year. The company has since doubled down and continues to release new products for voice AI and Slack at a rapid pace.

Salesforce credits its customers for the rate of its product releases. The company told TechCrunch that by letting its customers lead the way, it is able to build an AI product roadmap that can quickly react to where AI technology is headed.

When large language models were introduced, enterprises naturally wanted to jump on the technology but didn’t have the last-mile tech needed to fully use LLMs, Muralidhar Krishnaprasad, the president and chief technology officer of Salesforce engineering, told TechCrunch.

Techcrunch event Meet your next investor or portfolio startup at Disrupt

Your next round. Your next hire. Your next breakout opportunity. Find it at TechCrunch Disrupt 2026, where 10,000+ founders, investors, and tech leaders gather for three days of 250+ tactical sessions, powerful introductions, and market-defining innovation. Register now to save up to $410. Meet your next investor or portfolio startup at Disrupt

Your next round. Your next hire. Your next breakout opportunity. Find it at TechCrunch Disrupt 2026, where 10,000+ founders, investors, and tech leaders gather for three days of 250+ tactical sessions, powerful introductions, and market-defining innovation. Register now to save up to $410. San Francisco, CA | REGISTER NOW

The need for that last-mile tech is what sparked Salesforce to launch its agent management platform Agentforce, Govindarajan said.

From there, the company developed a bottom-up strategy led by themes — including agent context, observability, and deterministic controls, among others — as opposed to specific product timelines. This approach uses direct feedback from rotating groups of customers to build products with the assumption that other enterprises will have similar needs.

Customers in the driver’s seat

“The innovation that we’ve brought, they are direct result of us working with a vast number of these customers and then classifying the problems they see in the real world,” Govindarajan said. “Then [we break] that down and say, which of this can be solved at the LLM layer, which cannot? And for things that we cannot solve at the LLM layer, we need to build that sort of agentic operating system components around the LLMs to be able to go do that.”

Working so closely with customers’ engineering teams allows Salesforce to fix problems quickly before the technology evolves past them.

“We can’t wait three months or six months to get feedback, and then go figure out another six months of work,” Krishnaprasad said. “We are literally reacting to it, week by week, month by month. That’s been a big change. Now we push code, pretty fast, and we have various sorts of gates to try out new features, get earlier feedback before we release it broadly as well. So those are all changes that we had to do to kind of accommodate this rapid change in this environment.”

Engine, a travel management platform, is one of the companies within Salesforce’s customer feedback loop. And it’s not a casual relationship. The company’s operations team meets with Salesforce weekly, according to Engine founder and CEO Elia Wallen.

Through the partnership, Engine gets access to AI tools before they’re released. Wallen said the access helps Engine stay competitive and get more value out of these tools than it would otherwise.

The relationship goes both ways.

Wallen said he’s seen feedback from Engine get implemented into Salesforce tools. For example, Wallen said he instructed an AI voice agent to book him a hotel in Chicago. He thought the voice and interaction felt a bit unnatural and shared that with Salesforce. Shortly after, the agent had been changed and the company’s A/B tests started showing better results.

“If somebody is willing to actually help curate and build products that we need, they can help us better and really understand our problem and how they can solve it,” Wallen said. “For us, it’s fantastic to actually be invited into a thing like that, because we can influence the product.”

This strategy also allows the company to roll out solutions and workflows built by users to its broader customer base too.

Federal credit union PenFed has been able to slim down its tech stack by working closely with Salesforce, Shree Reddy, the company’s chief innovation officer and executive vice president, told TechCrunch.

“We invest our time, energy into the platforms that are more strategic, and we obviously spend a lot more time on this relationship,” Reddy said about Salesforce. “That investment has yielded good results in terms of strengthening that partnership that’s influencing each other, and what we see is the best value add mutually to both organizations.”

Reddy said PenFed developed an IT service management (ITSM) workflow on its own using existing tools and agents in Agentforce that worked well for the company. Salesforce was able to see that success and roll out the tool into the broader platform for other enterprises to use as well.

The downside to this approach is that it relies on the classic service sentiment that the customer is always right. Salesforce is hoping they are despite many enterprises still figuring out what role AI will play in their business, and many having yet to find value from the tech. As a result, they might not be the best source for long-term product development.

Plus, being willing to test and preview technology in beta now doesn’t necessarily translate to long-term usage habits or future software contracts either.

Be your own biggest user

The company also takes this bottom-up approach internally. Govindarajan said Salesforce employees are the biggest users of its AI tools.

The company also shifted labor and resources at the start of the AI boom. When ChatGPT was released, Salesforce moved around teams and resources to create a new AI team — a strategy the company has found successful during different innovation waves in the past, Krishnaprasad said.

“As the technology changes, we never know what’s going to come out a month from now,” he said. “We will adapt to it. And that’s what we did all of last year. If you think about it, agents weren’t even in terminology when you look back a year and a half ago. And then we had to go react to it. We had to go react to all the advances, and we had to react to our customers.”

Uber’s forthcoming luxury robotaxi service with Lucid Motors and Nuro is getting a fourth partner: Hertz.

The companies announced Thursday that Hertz will provide “day-to-day vehicle asset management, including charging, maintenance, repairs, cleaning, and depot staffing.” The service, announced last year, is supposed to launch by the end of 2026 in the San Francisco Bay Area, using Lucid’s Gravity SUVs and Nuro’s self-driving tech.

Hertz is handling this work through a newly established affiliate it’s calling Oro Mobility, which the rental company says will “provide integrated fleet management solutions across a range of mobility segments.”

“As the industry transitions from personally owned vehicles to commercially operated driver-led and autonomous fleets, Oro aims to fill a critical orchestration and operations gap,” the Hertz press release reads.

This is not the first time Hertz, which went through a bankruptcy restructuring process in 2020, has followed new mobility trends.

The company made a big splash in 2021 when it announced it was buying 100,000 EVs from Tesla, news that helped Elon Musk’s car company reach a $1 trillion valuation for the first time (and helped Hertz’s image as it emerged from bankruptcy). Hertz also announced plans in 2022 to buy up to 175,000 EVs from General Motors, and another 65,000 from Polestar.

None of those deals were ever fully realized, and Hertz started a fire sale of the EVs it had bought in early 2024. It did that in part because of higher-than-expected maintenance costs due to Uber drivers renting the EVs, and because Tesla slashed prices to stave off competition and boost sales.

Techcrunch event Meet your next investor or portfolio startup at Disrupt

Your next round. Your next hire. Your next breakout opportunity. Find it at TechCrunch Disrupt 2026, where 10,000+ founders, investors, and tech leaders gather for three days of 250+ tactical sessions, powerful introductions, and market-defining innovation. Register now to save up to $410. Meet your next investor or portfolio startup at Disrupt

Your next round. Your next hire. Your next breakout opportunity. Find it at TechCrunch Disrupt 2026, where 10,000+ founders, investors, and tech leaders gather for three days of 250+ tactical sessions, powerful introductions, and market-defining innovation. Register now to save up to $410. San Francisco, CA | REGISTER NOW

Starting up a fleet management and operations arm, though, should be closer to Hertz’s core competencies as a rental car giant. Competitors like Avis are already doing this kind of work for Waymo. And with robotaxi companies seemingly keen to use third parties to manage this piece of the puzzle, Hertz could build a decent business with Oro.

To wit, Hertz and Uber said Thursday that they will “explore expansion opportunities in 2027.” Uber has deals with dozens of autonomous vehicle companies around the world, and has plans to order at least 35,000 robotaxi-ready vehicles from Lucid Motors alone in the coming years. It’s starting with 10,000 Gravity SUVs, and recently announced plans to order another 25,000 EVs from Lucid Motors that will be based on its upcoming mid-sized platform. (Uber also now owns more than 11% of Lucid Motors as part of investments it has made alongside the vehicle orders.)

Elon Musk’s X is still trying to woo back advertisers — this time with a rebuilt, AI-powered ads platform that started to roll out on Thursday.

The tech company initially struggled to grow ad revenue in the first few years of Musk’s reign, leading it to shift its focus to other monetization channels, including AI and subscriptions. However, forecasts from eMarketer indicate that X’s ad business has been turning around as of late, with estimated ad revenue of $2.26 billion in 2025, rising to $2.46 billion in 2026. While that’s still half the size of Twitter’s 2021 ad business, it’s headed in the right direction again.

Now, X is hoping to juice that growth further with a new ad platform.

According to X, it has begun a “phased rollout” of the new platform, which it claims will have more modern “retrieval and ranking systems,” powered by AI. These changes are meant to make it easier for marketers to create targeted campaigns that they can control. AI will be used to enhance the campaigns, offering better results, more relevant ad placements, and precise targeting, according to X.

“Very few companies would have the ambition and technical courage to completely rebuild their entire advertising platform in such a short timeframe. This is classic X and xAI — bold, fast, and focused on building something substantially better for advertisers,” said Monique Pintarelli, head of global advertising at xAI, in a statement posted on X. “We are designing this new ad stack to enable more rapid and seamless integration of ongoing innovation. Advertisers can expect a smooth delivery of continuous improvements and a regular drop of new features as we keep pushing the platform forward.”

It’s not surprising that rebuilding X’s ad platform was a priority for the company after it merged with Musk’s xAI last year.

AI has contributed to revenue growth in ad businesses across the tech industry, as this week’s earnings have shown. Google, Meta, and others have been enjoying a “digital ad boom,” The New York Times noted this week, as AI systems have helped to automate the various aspects of marketing, from ad creation to targeting to measurement. This has also lowered the barrier for smaller businesses, giving them access to the same tools used by corporate giants, the report said.

OpenAI and Anthropic have been on the warpath lately against third-party efforts to train new AI models by prompting their publicly accessible chatbots and APIs, a process known as “distillation.”

That conversation has focused on Chinese firms using distillation to create open-weight models that are nearly as capable as U.S. offerings, but available at a much lower cost. However, tech workers have widely assumed that American labs use these techniques on each other to avoid falling behind competitors.

Now we know it’s true in at least one case: On the stand in a California federal court on Thursday, Elon Musk was asked if xAI has used distillation techniques on OpenAI models to train Grok, and he asserted it was a general practice among AI companies. Asked if that meant “yes,” he said, “Partly.”

Musk is in the process of suing OpenAI, CEO Sam Altman, and Greg Brockman, alleging they breached the original nonprofit mission for OpenAI by shifting the entity to a for-profit structure. That trial began this week, featuring testimony from the tech leader.

Musk’s admission is notable because distillation threatens AI giants by undermining the advantage they’ve built by investing in compute infrastructure. This allows other software makers to create models that are nearly as capable on the cheap. There’s no small amount of irony here, given the bending and alleged breaking of copyright rules by frontier labs in their search for sufficient data to train their models.

It’s no surprise that Musk’s xAI, which started in 2023, years after OpenAI, would try to learn from the then-leader in the field. It’s not clear that distillation is explicitly illegal, but rather may violate the terms of service companies set for the user of their products.

OpenAI, Anthropic, and Google have reportedly launched an initiative through the Frontier Model Forum to share information about how to combat distillation attempts from China. These typically involve systematic querying of models to understand their inner workings. To stop the efforts, frontier labs are working to prevent users from making suspicious mass queries.

OpenAI did not respond to a request for comment on Musk’s admission at press time.

Later in his testimony, Musk was asked about a claim he made last summer that xAI would soon be far beyond any company besides Google. In response, he ranked the world’s leading AI providers, saying Anthropic held the top spot, followed by OpenAI, Google, and Chinese open source models. He characterized xAI as a much smaller company with just a few hundred employees.

專攻AI大數據分析的老牌廠商SAS在年度大會SAS Innovate 2026上，宣布擴充自家分析平臺SAS Viya的代理型AI功能，主打2大新能力：內嵌於分析工作流程的AI助理系列SAS Viya Copilot、讓外部AI代理呼叫SAS分析能力的MCP Server。同時，他們也鎖定企業的AI代理需求，打造一款人人可用的AI代理加速器Agentic AI Accelerator，降低不同專業技能人員建置AI代理的門檻。

Copilot不是外掛聊天介面，而是嵌進工作流程的AI助理

SAS Viya Copilot的定位，是直接嵌入SAS Viya平臺的對話式AI助理，整合微軟一站式AI平臺Microsoft Foundry，讓資料科學家、開發者和業務分析師能在既有分析環境中，用自然語言完成資料探索、模型建置、儀表板生成和洞察解讀等任務，不需要切換到另一個獨立的聊天介面。

目前Copilot支援的功能涵蓋：自動生成SAS和Python程式碼、模型管線建議與可解釋性說明、以自然語言建立儀表板，以及AI輔助的案件和預警資訊生成。SAS強調，有別於其他靠通用LLM知識的AI助理，Viya Copilot是以產業資料模型、術語與最佳實務為基礎，能執行更貼近產業情境的分析任務。

SAS也推出2大產業專用Copilot，包括金融風險工作流程專用的資產負債管理（ALM）Copilot，以及協助臨床資料探索與病患族群分析的健康醫療Copilot。SAS也打算在2026年持續擴大應用範圍，比如金融犯罪防制、製造業供應鏈最佳化等場景。

建置MCP Server，外部LLM可直接呼叫SAS分析能力

SAS Viya MCP Server採用Model Context Protocol標準，把SAS的分析模型和決策能力作為客戶教的工具，企業使用者可從自選的LLM介面（如Claude）直接呼叫SAS的治理和分析能力，不需要複製邏輯或繞過企業管控機制。這也是大會現場Demo中，Claude能直接呼叫SAS Intelligent Decisioning完成補貨決策的技術基礎。

瞄準不同技術程度的團隊需求，推加速器讓人人都能建置AI代理

另一方面，SAS也發現一個企業痛點，就是要建置和部署AI代理，通常需要相當技術能力的工程團隊。為降低這個門檻，他們推出SAS Agentic AI Accelerator，提供一套預先整理好的程式碼、元件、介面和最佳實務框架，讓不懂程式碼的業務人員、或具備低程式碼能力的分析師，到有完整開發能力的工程師，都能在同一套治理框架內設計和部署AI代理。

目前，Accelerator已在GitHub上開放給現有SAS Viya用戶使用。

RAM：讓AI代理回答時有企業內部資料當依據

SAS Retrieval Agent Manager（RAM）解決的是另一個常見問題：AI代理在回應問題時，如果只依賴訓練資料，很容易產生與企業實際情況脫節的答案。

RAM以在RAG（檢索增強生成）框架為基礎，可將企業內部的非結構化資料，如政策文件、案件紀錄、作業程序等，轉化為AI代理可即時查詢的知識庫，讓代理在回應時，能以企業自家的最新資料為基礎，而非憑空推理。

目前RAM以獨立產品形式提供，SAS計畫後續將其整合進SAS Viya平臺，讓平臺上的所有AI代理和助理都能直接存取。