Equity · Sep 23, 2026 · 33m 19m saved
With Shardul Shah, Partner at Index Ventures
Public companies exposed to cybersecurity are worth close to a trillion dollars combined, by Shardul Shah's count, and he treats AI as the start of a new security category rather than another turn of the cycle.
Most investors backing AI security lead with a product. Shah told company boards to start with an insurance policy, and said any defense that still depends on a person approving each step will be outspent by attackers running swarms of software agents.
"If a human is in the loop on the defense side, compute or cost will drive an overwhelming force."
Shardul Shah, a partner at Index Ventures since 2008, spoke on Equity, TechCrunch's startup podcast. He led or co-led six consecutive rounds in Wiz, at the time the largest aggregate investment Index had ever made, before Google bought the cloud security company for $32 billion.
The full interview is covered here so you can skip it. 33 minutes of audio, 14 minutes of reading.
Here are the 14 arguments that matter.
Key Takeaways
Cyber is a permanent need, not a cycle, with close to a trillion dollars of public market value exposed to it
The first step for any board is cyber insurance, bought through a broker, before choosing a single vendor
He sits on the board of Coalition, which recently bought Allianz's cyber insurance business
Periodic security testing is ending: an outside audit graded from 1 to 10 does not fit a world of continuous attacks
A single person could soon run a swarm of agents against targets across the globe, so defense has to become machine against machine
Pattern matching is a fallacy, in his view, which leaves an investor relying on intuition
He is not a visionary, he said, and founders who want one should go to Vinod Khosla
Index does not set sector budgets for its $3.5B across seed, venture and growth funds
Wiz alumni are already founding companies, including a robotics start-up Index backed before it was incorporated
It is rational for markets to have irrational participants, which is his answer on nine-figure early checks
Pausing frontier AI would not stop the spending, because industry is still slow to put today's models to work
1. A Trillion Dollars Of Cyber
Rebecca Bellan opened by framing the moment: concern about AI safety and rogue agents has lifted cybersecurity stocks, and start-ups in the field are raising nine-figure checks. Shah was last on the show in March, when Google bought Wiz for $32 billion, and she asked about his other security bets since then.
Close to a trillion dollars of listed value depends on security
I think if you sum up the market caps of public companies it's it's close to a trillion dollars of valuation exposed to cyber companies and I think that speaks to the permanent resilient need for security that's only growing as there's more technology in the world.
Shardul Shah
He said the useful question is whether AI is a new category of security, not where cyber sits in a cycle. Bellan pushed back that a cycle is happening anyway. She pointed to Dario Amodei's letter about pacing the frontier and to AI researchers quitting their labs over fears of extinction events, much of which, she said, could show up as cyberattacks carried out by people using AI or by AI acting on its own.
The host's framing: a crisis and a boom at once
So, it's like AI is simultaneously creating this cybersecurity crisis and a cyber investment boom at the same time.
Rebecca Bellan
2. Remake Dr. Strangelove
Before answering on Wiz, Shah asked Bellan whether she had seen Stanley Kubrick's Dr. Strangelove. She had not. He described it as a satire about a psychotic general with the power to end the world.
He wants the film remade for this decade
And my call is for Christopher Nolan to remake Dr. Strangelove because we're back.
Shardul Shah
He then split security into two needs: protection from ourselves, and protection from adversaries who have a reason to attack. Ami, the co-founder and chief technology officer of Wiz, talked with him about managing both from the start of the business, and Shah said that framing is where the AI security category begins to be reshaped. He suggested a film night; Bellan said she would watch it.
3. Insurance Comes First
Bellan listed Index's recent security investments, including 7AI, which builds AI security agents, and Frame, which does AI-enabled human risk and security training, and asked how they fit the shift. Shah stepped back to what a company decision maker should do first.
Offload some risk before buying any product
I think the first thing to actually do and if any listener is on the board of any company of any size, my advice would be to try to offload some risk and the way to do that is through cyber insurance.
Shardul Shah
He disclosed that he sits on the board of Coalition, a cyber insurer, and said he was not talking his book, because a considered purchase goes through a broker who picks what suits the organization. After insurance come people, process and technology, and only at technology does the choice between incumbent vendors and start-ups arise. He said he is biased toward start-ups.
Incumbent products, in his words, are patched together
And the way I think about most incumbents is they're selling parachutes that are made of patches with kind of sketchy stitches.
Shardul Shah
A buyer also faces pitches from more than a thousand security vendors, he said, and it is hard to tell a real solution from one that does not work.
4. Insurance Is Sunscreen
Bellan said she had not known cyber insurance existed. Shah compared it to sunscreen: a basic, healthy habit that many people only pick up as adults.
Insurance could become compulsory
Eventually it may become compulsory or or even mandatory.
Shardul Shah
Asked who sells it, he named Coalition, which he said recently bought Allianz's cyber insurance business, and large insurers such as Beazley that cover companies of every size, industry and geography. Bellan added that CrowdStrike and Palo Alto Networks were doing well at the time of recording.
5. Platforms, Not Point Tools
Bellan then read out figures on Index itself. She put it to Shah that Index had recently raised money across three funds for about $3.5 billion of capital in total, and that the firm's stake in Wiz was about 12% and worth about $3.8 billion at the sale.
The host's figures, which he said were a little high
you had like what was it like a 12% stake worth about 3.8 billion if I if my homework is correct at the time of the sale.
Rebecca Bellan
"you're close a little shy but that's okay," Shah replied. On how Index picks security companies, he said the firm backs platforms.
Enterprises want platforms, not single-purpose tools
Ultimately we think enterprises want to partner with platforms not point solutions and these are businesses that have multiple use cases for multiple people and multiple product areas over time.
Shardul Shah
He named 7AI as the example, an AI-native company aiming to be "the foundational security company," and compared it to Wiz, which started as a platform in cloud security even though that category had existed for a decade.
Existing categories are also changing, he said. Identity management used to cover people only and now has to cover AI agents too, which behave differently and need new building blocks. He cited Frame, founded by Tal Shlomo after he left Wiz, as built for the new kind of risk posed by both human and non-human actors.
6. A Collection Of Artists
Shah said the thematic account made Index sound more deliberate than it is.
Index works as a group of individual investors
that said it might sound like we're really intentional and thoughtful and thematic about how we invest in security but the truth is as index we invest almost as a collection of artists right we have the seed fund a venture fund a growth fund and these are the three families that you allude to that add up to 3.5 billion dollars but it's the same group of individuals who are investing across the three and our objective is to be the first investor in category defining companies and that's the starting point.
Shardul Shah
Each partner's method, whether a theme, a view on a business model or most often a view on a founder, is personal, he said. Index does not open a new fund with a plan for how much goes to each sector. He called the process "more organic and a little bit more chaotic."
7. Wiz As A Talent Pool
Bellan asked whether Wiz would become what PayPal or Palantir became: a company whose alumni found companies in the same field. Shah said it already happened with Adallom, the Wiz founders' earlier company and the first on whose board he sat, about a decade ago. Armis was started by people who left Adallom.
The Wiz founders attract founders
Wiz is to your point somewhat similar because Assaf, Ami, Yinon, Roy, the founders of Wiz attract really ambitious entrepreneurial individuals and they celebrate their growth and impact in the next leg of their career and journey.
Shardul Shah
Index also backed Enigma, founded by Jonathan Jacobi, who was on the Wiz team and is now building a foundation model for robotics.
He agreed to invest before the company existed
when when we invested in his business, we actually shook hands before the company was incorporated, frankly, before we knew what they were going to do.
Shardul Shah
He said he never pictured himself as a robotics investor. On the commercial side, he said Dali Rajic, who ran go-to-market at Wiz, becoming OpenAI's chief revenue officer was "an absolute coup," and called him "the best go-to-market leader on the planet."
8. Intuition Over Zoom
Bellan asked what draws him to a founder. Shah first disputed her suggestion that knowing someone makes a connection. He told Jonathan Jacobi, on their first call, that he does not make investment decisions over Zoom. Because of a health matter in Shah's family, Jacobi flew to New York.
A 20-hour first meeting
Jonathan has got a really warm heart and so he jumped on the next flight to see me in New York and on Saturday we spent the day together from 6:00 a.m. to 2:00 a.m.
Shardul Shah
He described the result as a sensation that investors usually call intuition. Partners rationalize it for themselves and their partnership, which he said is useful as a test and dangerous as a habit.
Too much reasoning can talk an investor out of conviction
And I actually think that it's like it's super interesting and dangerous to rationalize because sometimes we're too smart and we talk ourselves out of really deep conviction.
Shardul Shah
9. No Patterns To Match
On when to double down, Shah said Index led or co-led six consecutive rounds of Wiz, at the time the largest aggregate investment the firm had made, and he warned against judging those decisions by the outcome.
Each follow-on was frightening
Right in the moment I did feel fear like it's really scary every two months to come back to your partnership and say I actually know more and want to invest more.
Shardul Shah
His partners asked what more he could know after two months, one of them December. He said Index has often made the wrong call because investors "underestimate the magnitude of the largest companies in the world." Bets that fail happen "more often than not" in venture, he said, which takes "a degree of either delusion or masochism or confidence."
He rejects the investor's most common shortcut
I actually think there are no patterns and pattern matching is a fallacy.
Shardul Shah
Bellan argued that he is seeing patterns whether he calls them that or not, and compared it to a reporter's news judgment.
10. Not A Visionary
Shah said most entrepreneurs may not want to work with him, and that is fine. Founders often want a backer who shares their view of the future, so they do not have to argue for it.
Founders who want a visionary should look elsewhere
And so if entrepreneurs want a visionary, and I mean this with great respect, they should go talk to Vinod Khosla who's who can see the future, right?
Shardul Shah
What he backs is the person, he said, because some of his best companies changed direction entirely. He said founders who need a believer in their thesis can rule him out early and save their time.
11. Continuous Security
Bellan cited a warning from more than 100 companies that AI-enabled attacks will become more common and more sophisticated, and a new AI model's reported ability to find and exploit known vulnerabilities on its own. She asked which existing security categories AI would destroy.
Periodic testing is ending
You know, I think we need to very quickly evolve to a world of continuous security.
Shardul Shah
His example was security testing. Public companies are required to be tested against security controls periodically: an outside agency sends a team, writes a report and gives a grade or a score from 1 to 10, and the board asks why. He said that world "does not exist going forward," and a field of new companies is emerging around external penetration testing, meaning simulated attacks run by outsiders.
His second point was about attackers: nation states, criminal groups and individual hackers, with political or financial motives. AI lets more of them launch more sophisticated attacks for less money.
One person, 10,000 agents
So I can soon imagine a single individual in a rogue state that launches a agent swarm of 10,000 to conduct ransomware attacks across the globe.
Shardul Shah
Defense, he said, has to become a machine-to-machine fight, and any security business built around a human approving each step needs to change. That is the thesis behind 7AI. He expects a new stack: security foundation models, defensive swarms of agents, and services that make attacks more expensive to carry out.
12. People Make It Defensible
Asked what makes an AI security start-up defensible when the underlying models keep improving, Shah said it starts with the people.
Everyone involved is being irrational
Most companies will not succeed. So, it's it's completely irrational to start a business. It's actually irrational to join a company, and it's probably even more irrational to invest in it.
Shardul Shah
"We're all mad here," Bellan said. Shah said he is drawn to founders who can hold a contradiction, such as a product that is easy to use and still highly sophisticated, because the best companies build platforms.
13. Not The Next Wiz
Bellan asked what the next Wiz looks like. Shah said he avoids the comparison, a lesson he learned from Datadog, where he still sits on the board.
Success made him a snob once
Datadog was it is a wonderful business. I still serve on the board, but it made me a bit of a snob.
Shardul Shah
He said he worries the same could happen after Wiz, and that he was in "the right place at the right time with an incredible group of people."
14. Irrational Checks
Bellan read out Index's recent early-stage checks and asked what makes the prices rational.
The host's list of nine-figure early rounds
So index invested in Instinct which that was a $250 million Series B. I think I'm just going to rattle some stuff off. That was at a $2.5 billion valuation for a very young company.
Rebecca Bellan
She also cited Simile's $200 million Series B, Enigma's $71 million seed round and 7AI's $130 million Series A. Shah answered with Simile, which builds AI simulations of people. The first check was a bet on the team. Then Index saw demand from large companies across countries and industries.
Demand on a scale he had seen once before
And I hadn't seen that amount of market demand since the beginnings of Wiz.
Shardul Shah
That demand turned into signed contracts, renewals and larger contracts in a short time, which triggered the next investment. On stage with Simile's founder at an event in New York, Shah asked how he thinks about raising money. The founder said he wants every customer in the world but not every investor, and looks for "an irrational amount of belief." Bellan finished the thought: irrational belief produces irrationally sized checks. "Exactly," Shah said.
His economics lesson
It's perfectly rational for there to be irrational participants in a market.
Shardul Shah
He said he studied economics at the University of Chicago.
Bonus Insights
A paused frontier would not stop the spending
Bellan's last question was about Dario Amodei's essay on pacing the frontier: if frontier AI development were slowed, would the assumption behind current prices, that models keep improving and demand keeps rising, fall apart?
I think that the technological advancement in the last 2 years has definitely surprised me. However, the diffusion of that technology into industry into the hands of consumers is lagging.
Shardul Shah
The open question, he said, would be whether irrational behavior spreads from one group of people to another and starts a bank run. Bellan called that topic "extremely anxiety inducing" and closed the interview.
The two met at an Index dinner the week before
Bellan and Shah had met a few days earlier at a dinner Index hosted. He called her one of the best dinner guests the firm has hosted, and they had debated optimism, skepticism and cynicism there. Shah has been at Index for almost 20 years.
LinkedIn is the best place to reach him
Shah said he checks it often and accepts connection requests from people who are specific about their questions and ideas.
Shah's bottom line is that AI has made attacks cheap enough that security can no longer be a periodic check run by people, and that large early checks are rational once a company shows the kind of demand he saw at Wiz and Simile.
Products, Companies & Tools Mentioned
Wiz (The cloud security company Index backed through six consecutive rounds; Google bought it for $32 billion)
Google (Wiz's acquirer)
7AI (AI security agents; Shah's example of an AI-native security platform, which Bellan said raised a $130 million Series A)
Frame (AI-enabled human risk and security training, founded by Tal Shlomo after he left Wiz)
Coalition (The cyber insurer on whose board Shah sits; he said it recently bought Allianz's cyber insurance business)
Allianz and Beazley (Insurers named in the cyber insurance discussion)
CrowdStrike and Palo Alto Networks (Cybersecurity stocks Bellan said were doing well at the time of recording)
Adallom and Armis (The Wiz founders' earlier company, and a company started by people who left it)
Enigma (Jonathan Jacobi's robotics foundation-model start-up, which Index backed before it was incorporated)
Datadog (Where Shah still sits on the board; the success he says made him a snob)
OpenAI (Where Wiz's former go-to-market head is now chief revenue officer)
Simile (AI simulations of people; Shah's case study for large early checks)
Instinct (A consumer AI agent company; Bellan cited a $250 million Series B at a $2.5 billion valuation)
Index Ventures (Shah's firm, investing through seed, venture and growth funds)
Books & Resources Mentioned
Dr. Strangelove – Stanley Kubrick (The satire Shah wants Christopher Nolan to remake)
We Must Pace the Frontier – Dario Amodei (The proposal behind Bellan's last question)
Watch the full interview
🟣 Apple Podcasts | 🔗 Episode page
If this was worth your time, send it to someone closer to the industry than you are.
Get the latest market chatter as it happens: