{"id":191615,"date":"2026-02-28T12:47:00","date_gmt":"2026-02-28T17:47:00","guid":{"rendered":"https:\/\/testing.news-you-need.com\/index.php\/2026\/02\/28\/20vc-x-saastr-this-week-anthropic-wiped-20b-off-cybersecurity-100k-deca-millionaires-are-coming-and-why-every-b2b-product-feels-dated-now\/"},"modified":"2026-02-28T13:05:10","modified_gmt":"2026-02-28T18:05:10","slug":"20vc-x-saastr-this-week-anthropic-wiped-20b-off-cybersecurity-100k-deca-millionaires-are-coming-and-why-every-b2b-product-feels-dated-now","status":"publish","type":"post","link":"https:\/\/testing.news-you-need.com\/index.php\/2026\/02\/28\/20vc-x-saastr-this-week-anthropic-wiped-20b-off-cybersecurity-100k-deca-millionaires-are-coming-and-why-every-b2b-product-feels-dated-now\/","title":{"rendered":"20VC x SaaStr This Week: Anthropic Wiped $20B Off Cybersecurity, 100K Deca-Millionaires Are Coming, and Why Every B2B Product Feels Dated Now"},"content":{"rendered":"

20VC x SaaStr This Week: Anthropic Wiped $20B Off Cybersecurity, 100K Deca-Millionaires Are Coming, and Why Every B2B Product Feels Dated Now<\/a><\/p>\n

https:\/\/www.saastr.com\/20vc-x-saastr-this-week-anthropic-wiped-20b-off-cybersecurity-100k-deca-millionaires-are-coming-and-why-every-b2b-product-feels-dated-now\/<\/a><\/p>\n

Publish Date: 2026-02-28 12:47:00<\/a><\/p>\n

Source Domain: www.saastr.com<\/a><\/p>\n

Author: <\/a><\/p>\n

Using an unordered list, summarize the following article with between 4 and 8 key points.
\n\t\t\t\t\tWith Harry Stebbings, Jason Lemkin, and Rory O\u2019Driscoll<\/p>\n

Anthropic wiped $20 billion off cybersecurity stocks with a single product release. The Citrini research piece predicting a \u201c2028 Global Intelligence Crisis\u201d sent shockwaves through the market. Figma posted an epic quarter at $1.2 billion ARR growing 40%\u2014and nobody cared because we\u2019ve all given up on the present. OpenAI leaked slides showing they need another $110 billion. And Jack Altman left his own $400 million fund to join Benchmark.
\nThe panic is overdone and real at the same time.
\nHere\u2019s the thing nobody is saying out loud: Almost all the B2B software we use today is terrible now. Not \u201ccould be better.\u201d Terrible. Because AI software has gotten so good that manually inputting data for 2 hours into your CRM feels like using a rotary phone. The leaders can\u2019t keep up with how dated their own products have become.
\nAnd the Fortnite circle keeps shrinking.<\/p>\n

\u00a0
\nKey Takeaways
\n1. The Anthropic Security Panic Was Months-Old News\u2014But the Overreaction Reveals Everything
\nAnthropic\u2019s security review feature wiped $20 billion off cybersecurity stocks. But here\u2019s what most people missed: these capabilities already existed. You could run an enterprise-grade security audit inside Replit using Claude Code weeks ago. The market panicked about features that have been live for months. The real story isn\u2019t the product\u2014it\u2019s what happens when stocks are priced for perfection. CrowdStrike at 16x revenues with 22% projected growth? Anything less than perfection is a kick in the nuts.
\n2. Only One Public B2B Company Has a Competitive Agent. It\u2019s Palantir.
\nNot one other publicly traded B2B company has seen a single ounce of revenue acceleration from their AI agents. Meanwhile, the startups we talk about each week have jaw-dropping acceleration. The difference? Incumbents are sprinkling AI dust on analytics software. Startups are building the agents that actually matter in their category. Hyper-niche agents work because they have a small set of things to do. Broad platforms like Monday or Shopify with 100 verticals can\u2019t build a specific agent that serves churches, basketball courts, and refrigerator businesses all at once.
\n3. Ghost GDP Is Already Here\u2014Jason\u2019s Team Went from 12 to 2 People
\nSaaStr generates eight figures in revenue with two humans and a fleet of AI agents. Those agents\u2014Repley, Art, Qualie, Monty\u2014generate millions in revenue. But they buy nothing. They don\u2019t buy handbags, shoes, Netflix subscriptions, or T-shirts. The only thing they purchase is tokens. This is Ghost GDP: productivity gains that don\u2019t flow to human workers who then spend in the economy. It\u2019s not theoretical. It\u2019s happening right now.
\n4. The Citrini Piece Is Clickbait\u2014But the Micro-Level Disruption Is Real and Accelerating
\nThe macro argument that civilization ends because B2B software gets disrupted is absurd. Software represents roughly 2% of GDP. The rest of the world would say \u201cI\u2019m willing to lose those guys.\u201d But at the micro level? The disruption is faster and better than anyone expected. Security audits from an airplane. Code review that makes a mediocre board engineer obsolete. Everything Anthropic ships this year is ahead of schedule. The real question isn\u2019t whether disruption happens\u2014it\u2019s whether it happens fast enough to create structural economic dislocation before workers can adjust.
\n5. Momentum Beats Value Right Now\u2014But Know What Game You\u2019re Playing
\nOnly five core B2B stocks are up over the trailing 12 months: Palantir, Figma, Cloudflare, Shopify, and a handful of others. Everything else is down\u2014some catastrophically (Atlassian down 74.85%, Klaviyo down 58%). In an age of maximum uncertainty, momentum investing has been the only play that works in both public and private markets. But momentum plays at 40-50x revenues carry blow-up risk. And value plays, while historically better over five years, have been dead money for 18 months. Choose your time horizon and commit.
\nAnthropic\u2019s $20 Billion Wake-Up Call for Cybersecurity
\nThe market reaction was dramatic: Cloudflare, CrowdStrike, and others got hammered. But Jason had a different take\u2014one informed by actually using the product on an airplane the week before.
\n\u201cI literally did this on the plane flying back last week. I ran a detailed security audit on my code inside Replit. It did everything\u2014static code review, penetration testing, the works. It can already do a better security audit than a mediocre board engineer will ever do.\u201d
\nThe team at Replit didn\u2019t even know their own security audit had gotten that good. Jason sent the results to their entire technical team and they responded: \u201cWe didn\u2019t even know it was this good yet.\u201d
\nSo the panic is about capabilities that already existed. Cloud Code reviewing Cobol code dropped IBM 10%\u2014but that feature had been in Claude for months.
\nRory provided the nuanced counterpoint: the capabilities are real, but the enterprise adoption path still runs through companies like CrowdStrike. \u201cSoftware is the means by which AI will diffuse into the enterprise,\u201d he quoted from an HSBC report. These code-scanning capabilities will reach enterprises through existing vendors, not through raw Claude access.
\nThe real problem? CrowdStrike was trading at 16x revenues on Friday after the initial hit. That\u2019s still priced for perfection. \u201cWhen you are priced for perfection, anything less than perfection will be a kick in the nuts,\u201d Rory explained. \u201cAny increase in tail risk of you not being the winner logically corrects pretty substantially.\u201d
\nThe Fortnite Analogy: Every B2B Company\u2019s Territory Is Shrinking
\nJason landed on the perfect metaphor for what\u2019s happening to incumbent software: Fortnite. The circle keeps shrinking. Claude keeps consuming more and more of your surface area. You\u2019re stuck on a smaller and smaller island where you need to own more and more market share just to survive.
\n\u201cI talked to three founders over the weekend of near-public companies. The advice I gave them: your agent is not great. You\u2019re being disrupted by the agentic layer.\u201d
\nHe sees it on LinkedIn every weekend. Leaders posting about adding AI to their email feature. Processing emails more efficiently. His response is blunt: you\u2019re going to become irrelevant in two years. Not because your customers will churn\u2014they\u2019ll renew. But your growth will fall so far that you become a zombie.
\nThe evidence is damning. Jason has two investments he made last year that he loved. Those products no longer have a reason to exist in their prior form because of Claude. They\u2019ve pivoted, they\u2019ve evolved, but the standalone use case evaporated.
\nAnd then last week, Claude Code launched the ability to preview apps directly inside the desktop. For many product people, you no longer need Replit or Lovable for that workflow. The circle shrank again.
\n\u201cThe flip side is if you nail the agent, look how much revenue these guys built. They built a billion dollars of revenue building the agent that didn\u2019t exist. If you can do something extremely high-value that could not be done before, you can close millions of revenue your first week. It\u2019s never happened before in the history of software.\u201d
\nWhy Public Incumbents Are Failing at Agents (And Startups Aren\u2019t)
\nThe observation is stark: of all publicly traded B2B companies, only Palantir has a competitive agent showing real revenue acceleration. Everyone else is treading water with AI features that don\u2019t move the needle.
\nJason identified two practical reasons:
\nFirst, agents are still essentially custom. Every agent needs to be trained. Every agent needs to be onboarded. Every agent needs its data cleansed. This is a massive amount of work for organizations that already think they\u2019re overworked. Institutional momentum is real and it\u2019s brutal.
\nSecond, the talent doesn\u2019t exist. You need forward-deployed engineers smart enough to train and deploy an agent. The average customer success person with a green-yellow-red dashboard cannot tune an agent. And the meta-challenge for Shopify, Monday, HubSpot, and Toast: you can\u2019t afford the humans to do it at scale.
\nHyper-niche agents work because they have a small set of things to do. As soon as you get to platforms like Monday with 100 verticals, it\u2019s nearly impossible to build a specific agent that automatically serves churches and basketball courts and refrigerator businesses.
\nThe incumbents\u2019 agents are in beta with six people using them. Maybe 60. \u201cThey\u2019re going to get killed by the startup that does it. Killed.\u201d
\nBut\u2014and this is the critical nuance\u2014it won\u2019t be the foundation model that sells directly to the church or the restaurant. There will be a software mediation layer, just as there was in traditional SaaS. The opportunity is building intelligence-led applications in specific verticals. That\u2019s where the next billion-dollar agents will come from.
\nFour Ways AI Gets to the Enterprise\u2014And Which Two Will Win
\nRory laid out the framework cleanly. Foundation model intelligence has to reach the enterprise. There are four paths:<\/p>\n

Enterprises buy directly from Claude. Everyone gets their software from the foundation model.
\nEnterprises build their own agents. Every company constructs its own AI layer.
\nExisting incumbents integrate AI. CrowdStrike, Shopify, Monday add intelligence to existing products.
\nNew startups build on top of foundation models. Companies founded post-2022 leveraging Claude, OpenAI, etc.<\/p>\n

His bet: paths 3 and 4. Enterprises won\u2019t build their own systems directly on Claude. And Claude isn\u2019t going to build focused vertical systems for every industry. There will be a software mediation layer.
\nJason\u2019s amendment: path 4 is where the explosive growth is. The startups he sees each week have jaw-dropping acceleration because they\u2019ve built the agents that actually matter in their niche. It\u2019s not about sprinkling AI dust on analytics dashboards. It\u2019s about building something so valuable that you can close millions in your first week.
\nGhost GDP: The Economic Problem Nobody Wants to Talk About
\nJason made it personal. SaaStr went from 12 employees to 2 humans plus AI agents. The business generates eight figures in annual revenue. The agents\u2014Repley, Art, Qualie, Monty\u2014create massive value. But they buy nothing except tokens.
\n\u201cOur 10 agents generate millions of revenue, but they buy nothing. They work all weekend long. They\u2019re good kids. They create a lot of noise. But they buy nothing. Nothing except tokens.\u201d
\nThis is the Ghost GDP problem from the Citrini piece, stripped of the clickbait: productivity gains that accrue to fewer and fewer people while reducing the number of consumers who can spend in the economy.
\nRory pushed back hard. Productivity gains have been the engine of growth for 200 years. We went from 80% of people on farms to 4%. We produce so much food we\u2019re all fat. The other 85% found other work. \u201cAcross the scope of history, productivity is freaking awesome. It\u2019s the only thing that\u2019s made us rich.\u201d
\nBut he conceded the short-term concern: if disruption happens extraordinarily quickly and people don\u2019t have time to adjust, you get structural dislocation. \u201cIf all the 45-year-old programmers are let go at the same time and 6 million programmers are on the street and there\u2019s no other work for them and it happens in a month, then yes\u2014in the short term there would be a GDP hit.\u201d
\nJason asked Claude to parse the data on a 50% reduction in tech headcount. The answer: $600 billion to $900 billion in GDP impact, 4 to 5 million total jobs lost including multiplier effects, and local economic devastation in five to six cities where tech is concentrated. \u201cIt would be one of the largest economic shocks in US history outside of a world war or pandemic.\u201d
\nBut Rory\u2019s reframe cut through the drama: even if the entire software industry got nuked, it would represent less than 1.5% of all US jobs. \u201cWe didn\u2019t bleed in Silicon Valley when the car industry went down the toilet. Don\u2019t hold your breath thinking they\u2019re going to come for us.\u201d
\nHarry pointed to Japan\u2019s 1990s as a recent precedent for productivity gains not dispersing to the broader population. Jason agreed\u2014meeting with Japanese B2B founders last November, every company was talking about how their seat base inherently shrinks each year as the economy contracts.
\nShopify\u2019s Quiet Warning Sign: We Haven\u2019t Needed That Many People for Years
\nToby L\u00fctke gets praised as one of the best CEOs in tech. Shopify is handling AI as well as any incumbent. And here\u2019s the number that should worry everyone: Shopify has the same number of employees it had three years ago. Not a single net headcount added. Revenue grew 50% to $12 billion.
\nThat sounds like efficiency. It is efficiency. But it\u2019s also an economic loss to the tech lifestyle we all lived just a few years ago. Revenue up 50% with zero headcount growth means all that incremental value goes to shareholders, not to new employees spending in the economy.
\nJason predicted something more dramatic: \u201cOne of the leaders in the next 12 months is going to do an Elon Musk and just cut half their team in one day. They\u2019re going to lay off half the entire company.\u201d
\nThe most likely candidates? The PE-backed, highly levered SaaS companies carrying 6x EBITDA in debt. If you bought a company at 8-9x and it\u2019s now trading at 4x with single-digit growth, the math doesn\u2019t solve any other way. You have to cut dramatically.
\nExpect to see five to eight startups at $50 to $200 million in revenue mashed together at nominal prices of 1.5-2x revenue. Frankenstein B2B companies with 20 products, professional management, and one shared ambition: maybe IPO in 2027.
\nFigma\u2019s Epic Quarter \u2026 That Nobody Celebrated
\nFigma posted Q4 2025 numbers that would have been cause for champagne two years ago: $1.2 billion ARR growing 40% year-over-year (accelerating from 38% in Q3), 97% GRR, 136% NDR among $10K+ customers. Stock up 15% after earnings.
\nRory framed it as the marquee fight: \u201cIn the right corner, Dylan Field, heavyweight champion of the world. In the left corner, Lovable, Replit, ringman Jason Lemkin.\u201d
\nJason\u2019s response was honest and conflicted. \u201cThere\u2019s nothing not to love in the quarter. Epic company. Epic quarter. We\u2019ve just given up on the present. We\u2019re all panicked about the future and you get no credit for a great quarter.\u201d
\nThe deeper concern: Figma\u2019s products are more workflow-oriented than people realize\u2014collaborative systems, not just pixel-perfect design tools. That gives them durability. But Jason would be shocked if by year-end, Claude Code can\u2019t automatically create designs as elegant as a professional designer\u2019s work.
\n\u201cIt has ingested every single website and mobile app on planet Earth. It can reproduce an iterative version that\u2019s just as good. The only reason it can\u2019t today is they haven\u2019t focused on it.\u201d
\nAnd Figma itself is leaning into the integration\u2014calling Claude Code integration one of their biggest growth drivers. But that\u2019s the Fortnite circle again. What happens when the native integration just overlaps more and more, and sometimes you skip Figma entirely?
\nThe Momentum vs. Value Debate: Where to Put Your Money
\nJason changed his mind since last week. Instead of bargain shopping, he\u2019s going momentum.
\nOnly five core stocks are up over the trailing 12 months: Palantir, Figma, Cloudflare, Shopify, and a small handful of others. \u201cIn an age of uncertainty, I\u2019m going to bet on whoever has gravitas. Momentum is gravitas because momentum builds on itself. This isn\u2019t fake hype. This is real momentum.\u201d
\nBut the tension is everywhere. The greatest dislocation in the market? Klaviyo versus Shopify. Klaviyo is essentially a Shopify derivative\u2014nearly 100% revenue attached. Yet Shopify is up 2.63% over 12 months while Klaviyo is down 58%.
\nAs a bargain hunter, you\u2019d buy Klaviyo. But Rory identified the trap: if Shopify is going to thrive, Toby has to build agents on top of his platform. And fundamentally, he has to take the market cap that currently sits with Klaviyo. \u201cHe probably has to kill you in order to survive.\u201d
\nThe most extreme dislocation: Palantir up 41% versus Atlassian down 74.85% over the same period. Atlassian is accelerating from 20% to 23% growth at $6.3 billion in revenue. It\u2019s the biggest decliner of the group despite accelerating. \u201cAs armchair value investors, you couldn\u2019t find anything better than Atlassian. Accelerating and beaten up.\u201d
\nRory provided the framework: over 6 to 18 months, momentum plays work and value plays don\u2019t. Over five years, value plays work and momentum\u2019s advantage disappears. \u201cThe trick is figuring out when you\u2019re transitioning from one to the other.\u201d
\nJason\u2019s decision: use a one-year lookback as the measurement point and buy from the winners. \u201cI\u2019ve already made the value bets in the past. I lost money on all of them.\u201d
\nOpenAI\u2019s $110 Billion Board Meeting
\nOpenAI is doubling spend to $665 billion by 2030 while upping revenue forecasts by 27% to $280 billion\u2014based largely on products that don\u2019t exist today: hardware, ads, and other new business lines.
\n\u201cWe\u2019ve all had this board meeting,\u201d Jason laughed. \u201cGood news and bad news. Good news is we\u2019re making up more revenue 3 years out. Bad news is I need another $110 billion to get there.\u201d
\nThe slides, leaked to The Information, felt like a startup board meeting on steroids. A beautiful stacked chart where three of the colors have never been done yet. \u201cWe\u2019re raising our forecast 30%. And we just need another $80 million to do it. Except with 3x zero attached to every number.\u201d
\nRory noted the narrative shift: Claude now gets the benefit of the doubt\u2014people believe it can kill everything. OpenAI, the former darling, gets no benefit of the doubt. But \u201cnobody is ever as good or as bad as they seem.\u201d
\nThe practical assessment: OpenAI is still the clear winner in consumer mind share. But when you run the math\u2014consumer subscription, enterprise (lagging Anthropic), agentic products that don\u2019t exist, and $30-77 billion in \u201cconsumer monetization beyond subscription\u201d\u2014there\u2019s a lot of leaning into the future.
\n\u201cAs long as these companies are perceived as so powerful that they can destroy everything, they will be able to get money. Because if you believe the models are going to take over the world, the only rational response as an investor is: I better get me some models.\u201d
\nQuotable Moments
\nJason Lemkin
\n\u201cAlmost all the B2B software we use today is terrible now. It\u2019s terrible. I can\u2019t talk to it. I can barely bring myself to use WordPress. All these products are terrible now because AI software is so good.\u201d
\n\u201cI talked to three founders over the weekend of near-public companies. Your agent is not great. You\u2019re being disrupted by the agentic layer. They are all being disrupted in real time and they are stressed as f.\u201d
\n\u201cOur 10 agents generate millions of revenue, but they buy nothing. They work all weekend long. Repley, Art, Qualie, Monty\u2014they\u2019re good kids. They create a lot of noise. But they buy nothing. Nothing except tokens.\u201d
\nHarry Stebbings
\n\u201cYou only have to have a partial deceleration to your numbers. You only have to have a partial risk. You only have to see more of the value of HubSpot or DocuSign flow to an agent for these stocks to do worse.\u201d
\n\u201cI have the data. I have an investment in this space. I already know the answer is yes. It\u2019s not my opinion. I have data from over 10,000 restaurants.\u201d
\n\u201cYouTube is the number one way we consume video and it is entirely based on a recommendation engine. It is the best recommendation company on planet Earth. Channels don\u2019t matter anymore. Followers don\u2019t matter anymore. Nothing matters.\u201d
\nRory O\u2019Driscoll
\n\u201cWhen you are priced for perfection, anything less than perfection will be a kick in the nuts.\u201d
\n\u201cAcross the arc of the last 200 years since the industrial revolution, productivity gains have been good. We used to have 80% of people working on farms. We now have 4%. And we sell so much food that we\u2019re all fat.\u201d
\n\u201cIf the only thing that\u2019s impacted here is the B2B software industry, my suspicion is the rest of the world will go, \u2018Yeah, I\u2019m willing to lose those guys.’\u201d<\/p>\n

<\/p>\n","protected":false},"excerpt":{"rendered":"

20VC x SaaStr This Week: Anthropic Wiped $20B Off Cybersecurity, 100K Deca-Millionaires Are Coming, and…<\/p>\n","protected":false},"author":1,"featured_media":191616,"comment_status":"closed","ping_status":"","sticky":false,"template":"","format":"standard","meta":{"fifu_image_url":"https:\/\/i0.wp.com\/www.saastr.com\/wp-content\/uploads\/2026\/02\/harry27.webp?fit=640%2C480&quality=65&ssl=1","fifu_image_alt":"","footnotes":""},"categories":[15],"tags":[26,24],"class_list":["post-191615","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-cybersecurity","tag-ai","tag-cybersecurity"],"_links":{"self":[{"href":"https:\/\/testing.news-you-need.com\/index.php\/wp-json\/wp\/v2\/posts\/191615"}],"collection":[{"href":"https:\/\/testing.news-you-need.com\/index.php\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/testing.news-you-need.com\/index.php\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/testing.news-you-need.com\/index.php\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/testing.news-you-need.com\/index.php\/wp-json\/wp\/v2\/comments?post=191615"}],"version-history":[{"count":1,"href":"https:\/\/testing.news-you-need.com\/index.php\/wp-json\/wp\/v2\/posts\/191615\/revisions"}],"predecessor-version":[{"id":191617,"href":"https:\/\/testing.news-you-need.com\/index.php\/wp-json\/wp\/v2\/posts\/191615\/revisions\/191617"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/testing.news-you-need.com\/index.php\/wp-json\/wp\/v2\/media\/191616"}],"wp:attachment":[{"href":"https:\/\/testing.news-you-need.com\/index.php\/wp-json\/wp\/v2\/media?parent=191615"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/testing.news-you-need.com\/index.php\/wp-json\/wp\/v2\/categories?post=191615"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/testing.news-you-need.com\/index.php\/wp-json\/wp\/v2\/tags?post=191615"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}