AI-native companies increasingly sell into a different reality. The buyer may test the product before speaking to anyone, run a proof of concept in a few days, and arrive at the first call with an informed view of the product’s quality, integration effort and likely ROI.

That changes the job of GTM. The goal is not simply to create more meetings. It is to identify real intent earlier, help a prospect reach value quickly, and turn product usage into a stronger sales and expansion motion. In many cases, the decisive moment is a 30-day pilot with clear success criteria, not a long sequence of human hand-offs.

Signals and Data Enrichment

The old model began with a list. The modern one begins with a signal: a relevant hire, new funding, a technology change, a visit from a target account, a sharp rise in product usage, or public evidence that a company has the problem you solve.

Clay is the best-known example of the new enrichment layer: it lets teams combine multiple data sources, web research and workflows rather than accept the limits of one database. Apollo remains a pragmatic option for contact data and basic outreach. Common Room is useful once you have enough channels and community or product activity to turn into actionable buyer signals.

The key idea is a data waterfall. Rather than relying on one provider for every email, job title or technology tag, query several in sequence and keep the best verified answer. That is useful. Buying every possible intent-data package before you know your triggers are not.

What this means for a founder: Choose three to five signals that reliably precede a good conversation. A list of 200 relevant accounts is more valuable than 2,000 generic names.

Outbound and Inbound Orchestration

This is the noisiest layer of the stack. Tools such as Apollo and Amplemarket make outreach more assistive: they help identify accounts, prepare research, draft messages and coordinate follow-up. Agentic offerings such as Artisan and 11x are aiming at a more autonomous model, from prospecting to meeting booking.

Both models can be useful, but they are not interchangeable. An autonomous agent works best when the audience, trigger, offer and message are already well understood. It is not a substitute for discovering them. Handing an unproven message to an agent merely automates your uncertainty.

Inbound is a better early use case for autonomy. Qualifying a form fill, answering routine questions, routing a lead, suggesting a meeting time and preparing an account brief are structured tasks with immediate feedback. The cost of delay is high and the brand risk is lower.

What this means for a founder: Let AI remove research and first-draft work. Keep a human close to strategic accounts, senior buyers and new messages until you know they earn a response.

Content, SEO and Distribution

AI has made publishing cheap. It has also made undifferentiated content close to worthless.

Every company can now produce a competent article, a week of social posts and a set of SEO pages. The output often looks polished but says nothing a buyer could not find elsewhere. That is not a content engine. It is a treadmill.

Use tools such as Claude for drafting and repurposing, Ahrefs for search research and AI-visibility monitoring, and Beehiiv or a similar platform when an owned audience matters. But begin with one original input: a customer insight, product benchmark, non-consensus opinion, data set, or story from the field. AI can multiply that input across formats; it cannot manufacture the underlying point of view.

What this means for a founder: Your moat is not publishing velocity. It is a perspective that compounds because your company sees something others do not.

Sales Engagement and CRM

Your CRM should be the company’s memory, not a reporting chore for a future sales team.

HubSpot is often the sensible default for an early sales-led motion because it gives a small team CRM, marketing and customer context in one place. Attio is compelling when the business does not fit a clean, linear pipeline: product-led expansion, partnerships, developer relationships, or usage-based accounts with multiple stakeholders.

The practical distinction is not “legacy” versus “AI-native.” It is whether the CRM contains enough trusted context for AI to do useful work. An agent cannot prioritise an account, produce a useful brief or recommend a next step when key information is scattered across Slack, email, product analytics and a founder’s memory.

What this means for a founder: Pick a simple source of truth early. Do not choose Salesforce because it looks grown-up, and do not customise any CRM until the motion is repeatable.

Customer Success, Expansion and Retention

For an AI company, customer success increasingly begins in the product data.

The most useful signals are rarely vanity metrics. Look for whether a customer has reached the high-value workflow, added teammates, expanded use cases, increased usage steadily, or hit a recurring point of friction. Those are the events that should trigger onboarding, support, a renewal conversation or a commercial expansion.

Vitally can help a lean team turn health scores into action. Intercom’s Fin is a strong example of where an AI agent can take on routine support, letting humans focus on difficult cases and strategic customers. But a health score is only useful if it changes a behaviour: alert an owner, trigger a playbook, or surface an expansion opportunity.

What this means for a founder: Instrument the journey to durable value before hiring a large customer-success function. Product usage is not a dashboard metric; it is a commercial signal.

Analytics and Attribution

Traditional attribution separated marketing, sales and product. That has always been an artificial division, and it is especially misleading when a buyer discovers you, tries the product, invites colleagues and only then starts a sales conversation.

PostHog is a sensible seed-stage choice because it brings product analytics, session replay, experimentation and feature flags closer together. Later, a warehouse stack may make sense. At the beginning, the goal is not perfect multi-touch attribution. It is being able to answer basic questions: which accounts activate, which actions predict conversion, where do pilots stall, and what usage predicts expansion?

What this means for a founder: Start with a data model, not a dashboard. What you instrument now determines which decisions you can make a year from now.

The Assistive-versus-Autonomous Line

The most useful debate is not whether agents replace SDRs. It is which work is structured enough for autonomy and which work still benefits from human judgment.

Full automation is already credible for enrichment, account research, inbound qualification, meeting scheduling, data hygiene, routine support and first drafts. These are repetitive workflows where speed matters and errors are relatively easy to catch.

Human judgment still wins in complex discovery, pricing, objection handling, strategic relationships and brand voice. The reason is not sentimentality about salespeople. It is that these moments require context, judgment and occasionally the ability to say something unexpected.

The hybrid model is therefore not an interim compromise. It is the best operating model for most seed and Series A companies: agents prepare, qualify, route and follow up; people make the consequential calls.

The Mistakes to Avoid

• Over-tooling before product-market fit. A 12-tool stack does not compensate for an unclear ICP, a weak proof point or an untested message.
• Buying “all in one” too soon. Consolidation is valuable after you understand your operating system. Before that, it can lock you into mediocre workflows and hide what is actually broken.
• Automating outbound without review. An agent can scale relevance, but it can also scale a stale message or an awkward tone directly into the inboxes that matter.
• Confusing activity with pipeline. More sequences, content and dashboards do not create demand. Relevance, timing, proof and follow-through do.

A Lean Stack for Seed Stage

Battery ventures published the modern AI native GTM landscape capturing the transitional move from manual labor to agentic AI.

For a seed-stage B2B AI company, I would keep the stack deliberately small:

• Clay + Apollo for enrichment, contact data and targeted outreach.
• HubSpot Starter or Attio as the account and relationship system of record.
• PostHog for the product signals that should shape acquisition, pilots and expansion.
• Claude, Search Console and Ahrefs for research, original-content production and measurement.
• Intercom with Fin or Vitally only when support or customer complexity is sufficient to justify a dedicated layer.

I would not add a fully autonomous AI SDR platform until the company has one repeatable outbound play, clear qualification criteria and a message that founders are proud to send themselves.

The Stack Will Keep Changing

Over the next few years, individual categories will blur. Enrichment will move into CRM, support agents will become onboarding and expansion agents, and product usage will become a more important GTM signal than a form fill.

The durable advantage will not be owning the newest tool. It will be building a system that connects signals, product value and human judgment faster than competitors can.

For founders building GTM infrastructure, AI agents or AI-native companies with an unusual path to market, Remagine Ventures would love to hear from you. We invest early in founders using AI to reimagine how people spend their time, attention and money.

According to the Israel Innovation Authority’s State of High-Tech 2026 report, Israeli high-tech generated approximately $84 billion in exits in 2025, alongside $85 billion in exports and nearly $15 billion in fundraising. High-tech output grew by 8.2%, and the sector accounted for 58% of Israel’s total exports.

Given that backdrop, the natural question entering 2026 was whether Israeli M&A would slow down after such an exceptional year.

So far, the answer is: not really.

The first half of the year has already produced a long list of strategic acquisitions involving Israeli startups. The buyers include some of the world’s most important technology companies: Apple, Nvidia, Cisco, Palo Alto Networks, CrowdStrike, ServiceNow, Motorola Solutions, Medtronic, PayPal, Akamai, SailPoint, 1Password, Western Union, Nuvei and others.

But the more interesting story is not just that global companies are still buying Israeli startups.

It is that Israeli M&A in 2026 is becoming more layered:

1. International acquirers are buying Israeli technology to fill strategic product gaps.
2. Multinationals that already have R&D activity in Israel are doubling down through additional acquisitions.
3. Israeli and Israeli-rooted companies are increasingly acquiring other Israeli startups.

That third category, the “blue-and-white” M&A story, may be one of the clearest signs of ecosystem maturity.

The biggest Israeli startup M&A deals of 2026 so far

The largest reported transactions in 2026 so far span fintech, defense tech, AI, semiconductors, robotics, medtech and cyber.

This is not 2021-style froth. The market is more selective. But that is precisely what makes the 2026 list interesting. Buyers are not simply acquiring growth stories. They are buying strategic capabilities.

What are acquirers buying?

The 2026 list shows clear demand for several categories where Israeli startups have historically been strong — and where global strategic buyers now have urgent needs.

The common thread is that acquirers are not buying “nice-to-have” products.

They are buying capabilities they believe they need in order to compete.

AI is not a sector. It is the acquisition rationale.

One of the clearest patterns in the 2026 M&A data is that “AI” is no longer just a category. It is the logic behind many transactions.

Some deals are obviously AI-related, such as Nvidia acquiring Illumex, ServiceNow buying Traceloop and Pyramid Analytics, and Apple acquiring Q.ai.

But AI also appears indirectly across the list.

In cybersecurity, AI is changing the threat model. The rise of agents, automated workflows and non-human identities is pushing buyers toward companies that can secure machine-to-machine interactions, enterprise browsers, data access, permissions and automated response.

That explains why companies like Cisco, Palo Alto Networks, SailPoint, 1Password, Akamai, Check Point, Cyera, Torq and Silverfort are all active around AI-era security surfaces.

In commerce, AI is changing how products are discovered and purchased. PayPal’s acquisition of Cymbio points toward a world where retail infrastructure needs to work inside AI assistants, search interfaces and emerging agentic commerce platforms.

In infrastructure, AI is changing the hardware stack. The acquisitions of DustPhotonics and Teramount point to the growing importance of photonics, optical connectivity and data movement in AI data centers.

AI is not one line in the table. It is the connective tissue across the table.

Cyber is still Israel’s main M&A engine

Cybersecurity remains the most obvious export category for Israeli technology.

The 2026 acquirer list includes:

The interesting thing is that this is not simply another year of classic cyber consolidation.

The categories are shifting toward:

• non-human identity
• AI-agent security
• browser security
• data governance
• automated response
• API security
• enterprise access control

That is a strong signal for founders and investors. The next cyber wave is being shaped by AI adoption, and Israeli startups are already being acquired in the emerging subcategories.

Defense tech and autonomy are now mainstream

The acquisition of D-Fend Solutions by Motorola Solutions for $1.5 billion is one of the most important deals of the year.

Counter-drone technology has moved from a specialist defense niche to a mainstream requirement for governments, airports, law enforcement, critical infrastructure and military organizations.

Other deals point in the same direction:

The convergence of drones, robotics, autonomy, defense software and AI is becoming one of the defining technology themes of the next decade.

Israel has a natural advantage in this area because of its defense ecosystem, engineering talent and operational urgency. But the category is no longer local or niche. It is global.

Semiconductors and photonics are having a moment

For years, Israeli tech was often described through the lens of software: cyber, SaaS, fintech, gaming, adtech and enterprise software.

The 2026 M&A list shows that the “harder” layers of technology are becoming strategic again.

This matters because AI infrastructure is not only about GPUs.

As AI clusters scale, data movement becomes a bottleneck. The physical movement of data between chips, servers and data centers is now a core part of the AI infrastructure stack.

That creates demand for optical connectivity, photonics, advanced semiconductors and networking infrastructure.

This also connects to one of the themes in the Israel Innovation Authority’s 2026 report: the renewed importance of deep tech, defense tech, space, quantum and AI infrastructure in the next generation of Israeli companies.

In a world where software moats are being compressed by AI, deep technical infrastructure may become more attractive again.

The “blue-and-white” M&A story

One of the most interesting patterns in the 2026 data is not just that global strategic buyers are active in Israel. It is that Israeli and Israeli-rooted companies are also becoming more active acquirers of Israeli startups.

This matters because it points to ecosystem maturity.

In earlier phases of the Startup Nation, the classic exit narrative was simple: build in Israel, sell to a US or European technology giant, become that company’s local R&D center.

That model still exists, and it remains important. But the 2026 data shows a more layered picture.

Israeli category leaders are now using M&A to consolidate talent, product capabilities and strategic wedges.

This is a healthy sign.

A mature ecosystem should not only produce startups that get acquired by foreign multinationals. It should also produce local champions that acquire teams, absorb technical capabilities and build larger platforms from within the ecosystem.

The rise of blue-and-white M&A also creates a local recycling mechanism. Founders, employees, capital and expertise do not simply leave the ecosystem after an exit. They often stay connected to it, either through the acquiring company’s Israeli operations or through the next generation of founders, operators and angel investors.

Are acquisitions still a route to multinational R&D in Israel?

Another question is whether foreign acquirers are using Israeli M&A to enter the market for the first time, or whether they are doubling down on existing Israel R&D operations.

The answer is both.

A public-source review of the 2026 acquirer list suggests that at least 40% of the international strategic acquirers already had a meaningful Israel R&D or technology operation before making their 2026 acquisition.

This group includes companies such as Cisco, Nvidia, Palo Alto Networks, Apple, PayPal, Medtronic, Akamai, ServiceNow, CrowdStrike, Nuvei, Ondas and Motorola Solutions.

That matters.

It means M&A is not only an entry mechanism. For many multinationals, Israel is already a strategic product and engineering base, and acquisitions are a way to keep feeding that base with new technology and talent.

At the same time, the list also includes buyers for whom the 2026 transaction appears to be a first or early Israeli foothold. 1Password’s acquisition of Apono was reported as its first acquisition in Israel and a basis for establishing a development center. Guardant Health’s acquisition of MetaSight was described as creating its first R&D center outside the United States.

That dual pattern is one of the defining features of Israeli M&A:

This is the real story of H1 2026. Israeli M&A is not only about exits. It is about where global technology companies choose to build, and where Israeli category leaders choose to deepen their advantage.

The macro context: strength, but also warning signs

The Israel Innovation Authority’s 2026 report gives important context for the M&A data.

On the positive side, Israeli high-tech returned to growth in 2025. The sector generated approximately $85 billion in exports, $84 billion in exits, and nearly $15 billion in fundraising. High-tech output grew by 8.2%, reached NIS 352 billion, and accounted for 18.3% of Israel’s GDP. The sector also accounted for 58% of Israel’s exports.

This is a remarkable level of resilience given the security, political and economic environment.

But the report also highlights structural concerns.

For the first time in more than a decade, the number of R&D employees in Israeli high-tech declined. The share of private Israeli high-tech company employees based in Israel has also continued to fall, from 69% in 2019 to 62% by March 2026.

That is the tension at the heart of Israeli tech today.

The ecosystem is producing valuable companies and attracting strategic buyers. But more activity, management and R&D are also moving abroad.

M&A can strengthen Israel as a global R&D hub when acquirers keep and grow local teams. But it can also accelerate the movement of decision-making abroad if local operations are not preserved and scaled.

That makes the blue-and-white M&A trend even more important. When Israeli companies acquire locally, the center of gravity is more likely to remain in the ecosystem.

What does this mean for founders?

For founders, the 2026 M&A data offers several practical lessons.

The companies being acquired in 2026 are not just “interesting startups”. Many own a strategic wedge. That is the bar in the current market.

What does this mean for investors?

For investors, the data reinforces several themes.

Cyber is still strong, but the next generation of cyber is shifting toward AI-era surfaces: non-human identity, agent security, browser control, data governance and automated response.

Deep tech is becoming more investable as the AI infrastructure buildout creates demand for semiconductors, photonics, robotics, autonomy and defense technologies.

Fintech is not dead, but the strongest exits are likely to come from infrastructure, compliance, payments, crypto rails, embedded finance and agentic commerce rather than consumer financial apps alone.

And perhaps most importantly: M&A remains a core path to liquidity in Israeli venture. The IPO window may reopen selectively, but the acquisition market is already active. Strategic buyers are still willing to pay for Israeli technology when it solves a strategic problem.

The bottom line

After a record-breaking 2025, 2026 is already showing that Israeli startup M&A remains active, strategic and globally relevant.

But the nature of the market is changing. This is not a broad, easy exit environment. It is a selective market where buyers are looking for specific capabilities: AI infrastructure, cyber for the agentic era, defense and autonomy, photonics, fintech rails, medtech, data and workflow automation.

The acquirer list is also a map of global demand. Apple, Nvidia, Cisco, Palo Alto Networks, CrowdStrike, Motorola Solutions, Medtronic, PayPal and others are showing what they believe matters. At the same time, Israeli companies like Mobileye, Fireblocks, Cyera, Torq, Check Point, eToro and Elbit are showing that Israel is not only a source of acquisition targets. It is also producing acquirers.

That is the more important signal. Israeli M&A in 2026 is not just about exits.

It is about the next structure of the ecosystem: foreign multinationals doubling down, new buyers entering Israel, and local champions consolidating from within.

One of the most repeated pieces of advice in venture capital is: own enough for it to matter. The logic is sound. Venture returns follow a power law, and if a fund does not own enough of its winners, even great investments may not return the fund.

That is why fund construction models exist. They force discipline around check size, ownership, dilution, reserves, and concentration. For emerging managers in particular, the model can feel almost sacred: write the right check, own the right percentage, and follow on in the companies that break out.

But like many useful rules in venture, this one becomes dangerous when applied too rigidly. Sometimes the question is not whether you can own 8 percent or 10 percent. Sometimes the question is whether you can get into the round at all.

What AngelList’s data suggests

AngelList recently published an analysis of more than 15,000 seed investments, and one of the most interesting findings was counterintuitive: smaller checks often performed better than typical-sized checks. That does not mean small checks are inherently better. It probably means something more nuanced.

Big caveat, the latest deals considered were from 2022, so the conclusions are based on ‘pre-AI’ data (pre-chatgpt and genAI craze, that is).

From the data, two key takeaways stand out:

• Small checks were the best performers. Despite being the smallest category, they delivered the highest average returns (outperforming typical checks by 1.17x) and the strongest 75th percentile returns—even though they had slightly more “losers”.
• Large checks showed slight outperformance. On average, large checks beat typical-sized checks by just 1%, making the signal less clear.

Angelist

GPs rarely choose to write small checks into companies they do not like. More often, they are forced into smaller allocations because the round is competitive and demand is high. In that sense, a small allocation can be a signal of quality rather than a sign of low conviction.

This matters because venture is becoming even more concentrated. More capital is flowing into fewer companies, and the most attractive rounds often get crowded very quickly. In a market like that, access itself becomes a source of edge.

Capital concentration changes the game

The current venture market is not evenly distributed. LP dollars are concentrating into fewer managers, VC dollars are concentrating into fewer startups, and later-stage capital is increasingly reserved for companies that already look like category leaders. The result is a sharper divide between the companies everyone wants to fund and the companies struggling to get noticed.

For seed investors, this creates a paradox. The model says you need meaningful ownership to generate fund-level returns, but the best companies may not give you the ownership your model requires. If you insist on perfect ownership every time, you may end up with a cleaner portfolio but miss the most important companies.

This is especially true in smaller ecosystems like Israel, where repeat founders, elite technical teams, and trusted insider networks can close rounds before the wider market even sees them. By the time a company is obviously hot, the allocation may already be gone.

Why a small check can still matter

The obvious objection is that small checks do not move the needle. In many cases, that is true. A tiny position in a good company may produce a nice mark-up, but it may not return the fund.

Still, a small check can create value beyond the initial ownership. It gives the fund a relationship with an exceptional founder, a seat inside the information flow, and potentially the right to participate later if the company breaks out. It can also create network value with co-investors, operators, future founders, and LPs.

In venture, being close to the best founders is itself an asset. The first check may be small, but the relationship can compound across future rounds, future companies, and future opportunities.

The model is a tool, not a religion

This does not mean funds should abandon ownership discipline. A portfolio full of tiny checks can quickly become a collection of logos rather than a strategy. LPs do not pay managers to collect access for its own sake. They care about fund-level performance.

But the model should not become a reason to miss exceptional opportunities. The job of the investor is to know when a small allocation is a compromise and when it is a privilege. There is a big difference between writing a small check because you lack conviction and writing a small check because the company is so competitive that this is the only way in.

A useful approach is to separate the core portfolio from an access sleeve. The core portfolio is where the fund sticks to its ownership targets, does the work, and builds concentrated exposure. The access sleeve is reserved for rare cases where the company is exceptional, allocation is scarce, and the answer to “why are we lucky to be invited?” is obvious.

Small tickets need a high bar

Small checks should not have a lower bar. They should have a higher one. The round should be genuinely competitive, the founders should be exceptional, the company should fit the fund’s thesis, and the co-investors should be people the fund respects.

The check should also be capped so it does not distract from the core strategy. A small access investment should not require the same time commitment as a lead position, and it should not become an excuse to avoid making harder, higher-conviction bets.

The best version of this strategy is not spray and pray. It is disciplined access. It recognises that ownership matters, but also that in a power-law business, the cost of missing the right company can be much higher than the cost of owning less than planned.

Sometimes the small check is the right check

Venture capital loves models because they make an uncertain business feel more controllable. They are useful, but they are not reality. The best companies rarely fit neatly into the spreadsheet.

Sometimes the model says the check is too small. The market is saying the round is oversubscribed, the founders are exceptional, and you are lucky to be invited. In those moments, the right answer may not be to pass because the ownership is imperfect.

Sometimes, getting in is the edge.

The original version of Firgun was published on Linkedin. Subscribe to get the next one in your inbox.

Firgun (pronounced ‘feer-goon’) is an Israeli term describing genuine, unselfish joy in someone else’s success… It’s basically the opposite of “schadenfreude”. Every week since March 2020, we celebrate the wins of our startup community. Read more about why Firgun matters. If this was forwarded to you and you’d like to subscribe, previous editions are on VC Cafe.

This week started with rockets from Iran, continued with a number of mega rounds for Israeli tech and finished with the 2026 FIFA world cup kick-off match. I spent the week in Tel Aviv and I can honestly say, Israel has a unique energy (especially in its tech sector). It’s vibrant, honest, hardworking and ambitious. I’ve enjoyed meeting founders in stealth and spending time with Remagine Ventures portfolio companies.

In the wider tech world, the defining story this week was Anthropic’s release of Claude Fable 5, a public Mythos-class model built for demanding reasoning and long-horizon agentic work, alongside the more restricted Claude Mythos 5 for approved customers. Anthropic faced backlash for initially applying hidden safeguards and rerouting some sensitive requests without making that clear to users, later admitting it had “made the wrong tradeoff.”. According to filings, Google has agreed to pay SpaceX $920M per month for access to Nvidia chips as part of a cloud-services deal that runs through mid-2029Apple used WWDC to put more substance behind Apple Intelligence; Google pushed new Gemini and open-model updates; and OpenAI reportedly continued moving toward an IPO while partnering with Visa to make agentic commerce real. Jeff Bezos’ Prometheus raised a massive $12 billion Series B at a $41 billion valuation to build what it calls an “artificial general engineer” for the physical world. OpenAI acquired Ona, a startup providing secure cloud environments for AI agents, to strengthen its Codex platform for long-running enterprise agents. KKR launched Helix Digital, a new vehicle to finance AI infrastructure with over $10 billion in committed capital, backed by NVIDIA and others. Anthropic signed multiple direct data center lease agreements as it continues to secure its own compute capacity. Other notable moves included PhysicsX raising $300M series C for aI infra to accelerate physical simulations for industrial clients, Genspark raising $100 million in a Series B extension at a $2.6 billion post-money valuation for its agentic workplace platform. The week underscored both the continued appetite for very large AI bets and the accelerating push toward production-grade agent systems and supporting infrastructure.

QUOTE OF THE WEEK

“Success is no accident. It is hard work, perseverance, learning, studying, sacrifice and most of all, love of what you are doing or learning to do.” — Pelé

NEW FUNDING ROUNDS

• Now it’s official: congratulations Yotam Segev and team Cyera on securing a $600M series G at a $12 billion valuation (after securing $400M in January…) for your AI-Native data security platform!
• Well done Tomer Aharoni and team Rylo on announcing your $85M series B to develop AI-powered real-time speech and sign-language translation tools for deaf and hard-of-hearing users!
• Way to go Alon Arvatz and team PointFive on your impressive $60M series B to help companies uncover and reduce wasteful cloud and AI infrastructure spending!
• Mabruk Yossi Torati and team A Security on coming out of stealth with a $37M round for your offensive security and remediation platform to protect against weaponised AI!
• Great stuff Ron Arbel and team Aryon Security on your $25M series A for your Cloud Security Enforcement Platform for pro-active cloud security!
• Congrats Guy Arazi and team Pi Security on coming out of stealth with a $25M series A and previously undisclosed $10M seed to analyse code, cloud infrastructure and workplace communications to identify real threats!
• Kudos Assaf Henkin and team Jedify on your $24M series A to build the context layer for enterprise AI!
• Big thumbs up Ido Bronstein and team Upriver on securing a $10M seed round (and previously undisclosed $4M pre-seed) to develop AI agents that autonomously manage enterprise data!

EXITS

• No new exits were recorded this week

LINKS FOR YOUR BROWSER

VC CAFE/ REMAGINE VENTURES/ ISRAEL VC

• Israel’s Space-Tech Moment: From Beresheet to Space-Based AI – Space is no longer just about rockets. It is becoming infrastructure for AI, defence, climate intelligence, communications, mining, science and the next generation of computing. Israel has a real shot to play a role.
• Agentic AI: Picks and Shovels vs. the Gold Itself – Agentic AI is really two very different bets: infrastructure and vertical applications. A Remagine Ventures take on where the value sits, and why token costs, non-determinism, and security will decide who survives.
• Quantum Computing’s Venture Moment: Is Now the Right Time to Invest? – Why Israel could become one of the important startup hubs in the next computing race

ISRAEL

• 2026 H1 Israeli startup Insights report by Bank Hapoalim ??? ??????? and Dealigence – Startup funding rose 24% to $8.6B+ and total exits reached $10.5B+. in H1 2026 – there’s also been a 2.1% growth in startup employment, driven largely by sales positions and an increasing shift toward offshore hiring (75% of new jobs).
• Why I Won’t Debate Critics of Israel A note to the Making Sense Community – by Sam Harris
• 12 moonshots for Israel -an interesting new book by Judah Taub

GLOBAL

• The 2026 Deep Tech Report – data rich report on the global state of deep tech by Drumbeat Capital
• The 2026 Tech Nation Report – a overview of UK/European tech and AI by Tech Nation
• The future of work and AI – the The Wall Street Journal interviewed economists on the impact of AI on workers, productivity and more.

MEDIA OF THE WEEK

Israel: with shockingly few people in the U.S., Israel is the #2 origin in total for foreign Unicorn founders (source)

Interesting graphic shared by former UK prime minister Rishi Sunak – Tel Aviv ranks #4 globally in terms of AI talent density… (source)

The fraud prevention market map by CB Insights

The web games market map

Who are the Israeli and international VCs investing in Agentic AI Israeli startups in 2026? infographic by Israel VC (source)

Who is funding Israeli Quantum computing startups – another great infographic by Israel VC (learn more about the Israeli quantum ecosystem on VC Cafe )

I miss proper Apple product launches. Siri AI is the latest in this time honoured tradition

That’s all for this week. Have a great weekend and Shabbat Shalom to the Firgun community. There’s a lot to celebrate in our “small” community! Keep on creating!

Eze Vidr

For most of the last decade, the reaction from generalist investors was predictable: fascinating science, huge market, impossible timelines. Quantum was the sort of thing that governments, universities, corporate labs and patient deep tech funds could afford to care about. Traditional VC, especially early-stage VC, largely stayed on the sidelines.

That may be starting to change.

In the past 18 months, quantum has moved from “interesting research frontier” to a more investable strategic category. The change is not because a useful, fault-tolerant quantum computer has suddenly arrived. It hasn’t. The change is that the industry is starting to look less like a collection of lab experiments and more like an emerging stack: hardware, control systems, error correction, compilers, quantum software, security, sensing, networking, cloud access, and eventually applications.

The latest signal came from Europe. In one week only, four quantum computing startups raised a number of mega rounds: Oxford Quantum Circuits (OQC) raised a $350 million Series C, France’s Quobly announced a €115 million Series A, and Finland’s IQM upsized its PIPE financing to $146 million ahead of a public-market transaction. The headline number in some reports was close to $1 billion of quantum funding in a week, though that should be treated carefully, as some coverage appears to count OQC and Oxford Quantum Circuits separately when they are the same company.

Even with that caveat, the direction of travel is clear: quantum is attracting serious capital again.

At the same time, McKinsey’s 2026 Quantum Technology Monitor argues that quantum has reached a “commercial tipping point,” with more than 300 organizations globally engaging with quantum computing and over $1 billion in quantum-computing revenue generated worldwide in 2025. The U.S., France, the U.K., the EU and China are all treating quantum as a strategic technology. Public funding, defence procurement, national security concerns and private capital are converging.

The question for venture investors is no longer whether quantum is scientifically important. It is whether now is the right time to invest.

The venture case for quantum

The strongest argument for quantum is simple: if it works, it changes what is computationally possible.

Quantum computers are not “faster laptops.” They are a different model of computation. Their promise is to solve certain classes of problems that are extremely difficult, expensive or impossible for classical computers to solve efficiently. That could matter in drug discovery, chemistry, materials science, logistics, financial risk, machine learning, cryptography and national security.

That “if it works” is doing a lot of work.

Quantum remains technically hard. Error rates are high. Scaling is difficult. Timelines are uncertain. Hardware approaches are fragmented: superconducting qubits, trapped ions, neutral atoms, photonics, silicon spin qubits, topological qubits and others. It is still unclear which architectures will win, or whether the future will be heterogeneous.

But this is also why the venture opportunity exists.

When a technology is obvious, derisked and close to revenue, it is usually too late for venture-scale returns. The role of venture is to fund the uncomfortable middle: after the science has become credible, but before the market has fully formed.

Quantum may now be entering that middle.

Several things have changed:

First, the technical roadmaps are getting more credible. Google’s Willow chip showed progress in quantum error correction. IBM has published a path toward fault-tolerant systems by the end of the decade. Microsoft, despite continued scientific debate around its topological approach, is also pushing toward a 2029 target. Startups such as Quantinuum, IonQ, PsiQuantum, IQM, OQC, QuEra, Alice & Bob and others are increasingly judged not only on papers, but on systems, customers, deployments and integration.

Second, quantum is becoming part of the AI infrastructure conversation. AI has reminded the market that compute is not a solved problem. As AI models become more expensive, energy-intensive and constrained by hardware bottlenecks, the search for new computing paradigms becomes more urgent. Quantum will not replace GPUs, but hybrid quantum-classical systems could become part of the future high-performance computing stack.

Third, governments are behaving like strategic customers. Quantum has implications for national security, encryption, defence, sensing and sovereign computing capacity. This matters for venture because government capital can extend runway, validate roadmaps and create early demand in markets that are too early for normal enterprise budgets.

Fourth, the investable surface area is expanding. Investors no longer need to bet only on “the company that builds the quantum computer.” There are opportunities in picks and shovels: control systems, cryogenics, lasers, photonics, chip design, error mitigation, software abstraction layers, compilers, developer tools, post-quantum security and vertical applications.

That is where Israel becomes interesting.

Why Israel is a good place to look for Quantum

Israel is unlikely to outspend the U.S., China or Europe on national quantum infrastructure. But Israel has rarely won by outspending larger markets. It wins when deep technical talent, military-grade problem solving, strong academia, entrepreneurial urgency and global capital collide.

Quantum fits that pattern.

The Israeli quantum ecosystem is still small, but it is unusually dense. A 2025 map by Earth & Beyond Ventures and Deloitte Catalyst identified nine core Israeli quantum-computing startups that had collectively raised around $650 million, including more than $300 million raised in 2025 alone. That is a meaningful amount of capital for a country of Israel’s size.

Israel has already produced globally relevant companies in quantum control, software, photonics, trapped ions, error mitigation and neutral atoms. The ecosystem has academic depth, national urgency, engineering talent and a growing investor base. It also has a history of turning constraints into advantage. The emerging Israeli stack includes:

Quantum Machines – building control and orchestration systems for quantum processors. This is a classic Israeli wedge: not necessarily owning the whole machine, but controlling a critical layer of the stack.

Classiq – building quantum software and tools that help developers design quantum algorithms and applications without needing to work at the hardware level.

Quantum Source – developing photonic quantum computers, a route that could potentially offer advantages in scaling and data-center integration.

Quantum Art – building trapped-ion quantum systems, with ambitions to deliver full-stack quantum computing.

Qedma – focused on error suppression and error mitigation, one of the hardest and most commercially relevant problems in making quantum useful.

QuamCore – working on scalable superconducting quantum architecture.

Q-Factor – a newer entrant building neutral-atom quantum computing, backed by investors including NFX, TPY Capital, Intel Capital and others.

Quantum Transistors – pursuing diamond-based and photonic approaches to quantum processors.

This matters because Israel is not producing only one kind of quantum company. The ecosystem spans hardware, software, photonics, control, error correction and architecture. That diversity is important because, at this stage, nobody knows which approach will dominate.

Israel also has several structural advantages.

1. Academic depth. Quantum companies are often born out of physics, electrical engineering, computer science and materials science departments, not from weekend hackathons. Israel has strong research institutions, including the Technion, Weizmann Institute, Hebrew University, Tel Aviv University and Bar-Ilan University, with a history of turning deep science into startups.
2. National security relevance. Quantum sensing, secure communications, cryptography and high-performance computing are not abstract commercial categories. They are strategic capabilities. Israel’s defense and intelligence ecosystem has historically created talent and urgency in areas where technical advantage matters.
3. Israel’s semiconductor and systems heritage. Quantum is not just physics. It requires electronics, control systems, software, materials, packaging, lasers, cryogenics and manufacturing know-how. Israel has decades of experience in chips, communications, cyber and systems engineering. That combination is useful in quantum, where the bottleneck is increasingly shifting from theory to engineering.
4. Founder mindset. Quantum founders need scientific credibility, but they also need commercial aggression. The best Israeli startups tend to be good at turning technical depth into global go-to-market motion. That will be critical in quantum, where early customers may be in the U.S., Europe, Japan and government-backed research programs.
5. The presence of specialist capital.

A small ecosystem can become powerful when it has repeat founders, repeat investors and repeat technical talent. Israel’s quantum market is not there yet, but the ingredients are visible.

What should investors look for?

For early-stage investors, the most interesting quantum companies will likely share a few traits:

• A wedge into today’s market. The company should not depend entirely on a 2035 quantum breakthrough. It should have a path to sell something useful earlier: control systems, developer tools, simulation, security, benchmarking, optimization, sensing, or infrastructure for quantum labs and enterprises.
• A credible technical founder-market fit. In quantum, technical depth is not optional. The team needs to be world-class scientifically, but also capable of productizing complexity.
• A clear architecture bet. Founders should be able to explain why their approach can scale, what the key technical bottlenecks are, and how they will prove progress over time.
• A non-dilutive capital strategy. The best quantum startups will often combine venture money with government grants, research partnerships and strategic capital. That is not a weakness. In deep tech, it can be a feature.
• A commercial bridge. Even if the big market is years away, there should be a path to near-term engagement with customers: pilots, paid research, hardware access, cloud deployments, licensing or strategic partnerships.
• A defensible layer. In AI software, moats can erode quickly. In quantum, defensibility can come from patents, know-how, hardware integration, scientific talent, manufacturing process, data, partnerships and long technical roadmaps.

Quantum will not be a quick flip. It will require patient capital, technical humility and a willingness to back teams before the market is fully legible.

But that is exactly where venture is supposed to operate.

The next computing platform will not be built overnight. It may not even look like the platforms we know today. But if quantum becomes part of the future compute stack, the best time to build positions may be before the category becomes obvious.

Is now the right time?

Quantum is not a normal SaaS category. It does not fit neatly into the standard pre-seed checklist of MVP, early users, fast iteration and short sales cycles. Like many other deeptech investments, the best quantum companies may require longer timelines, more non-dilutive capital, deeper technical diligence and more patience than the average venture-backed startup.

That makes quantum a poor fit for tourist capital. But it can be a strong fit for investors who understand three things.

First, the market will not arrive all at once. The first venture-scale returns may not come from the final universal quantum computer. They may come from enabling layers: software tools, control systems, integration platforms, error mitigation, simulation, quantum-safe security, sensing and hybrid quantum-classical workflows.

Second, the customer set is different. Early customers may be governments, national labs, cloud providers, pharma companies, banks, defense contractors, semiconductor companies and industrial giants. Selling into that world requires credibility, partnerships and patience.

Third, the timing of infrastructure markets is strange. The best time to invest is usually before the market is obvious, but after the enabling conditions are in place. In quantum, those enabling conditions are increasingly visible: large public programs, more credible technical roadmaps, serious private rounds, early enterprise budgets, and growing geopolitical urgency.

The venture question is not “will quantum be huge?” It probably will. The better question is: “Which parts of the quantum stack can become valuable companies before fault-tolerant quantum computing is fully mature?”

At Remagine Ventures, we’ve been mapping this space closely, including through IsraelVC, where we recently charted the named funds behind ten Israeli agentic AI startups that raised in the last few months (more on that below). The exercise crystallised something we’d been feeling in deal flow for a while: “agentic AI” is not one category. It’s at least two, and they have almost nothing in common as investments.

Infrastructure vs. Vertical: two very different bets

Agent infrastructure is the plumbing: runtimes, sandboxes, orchestration layers, identity and permissions, observability, evaluation, and security/control planes that make it possible to run an agent in a production environment without it doing something catastrophic. This is the “picks and shovels” layer. It’s technically hard, often defensible through deep systems work, and it sells to other AI builders as much as to enterprises.

Vertical AI agents are the applications: agents that do a job, like triaging support tickets, running procurement workflows, doing compliance review, managing revenue operations, or writing code in a specific framework. This is where the outcomes live, and increasingly where the capital is concentrating. Vertical agents account for roughly half of 2026 agentic AI deal volume and an even larger share of dollars deployed, with the busiest verticals being cybersecurity, healthcare operations, finance ops, procurement, compliance, and customer-facing support.

The infrastructure layer is smaller in capital terms but disproportionately active. Investors are chasing the execution and control layers that make agents safe enough to deploy in regulated or high-stakes environments. That’s a tell: the market has moved from “can we build an agent” to “can we trust an agent,” and that question is an infrastructure problem before it’s a product problem.

For founders, this distinction matters enormously when you’re raising. Infrastructure pitches need to prove technical depth and a path to being the standard that vertical builders depend on. Vertical pitches need to prove workflow ownership, data moats, and that you’re not just a thin wrapper that the underlying model provider could absorb in a quarter. Investors have sharpened their filters accordingly: proprietary data, novel architecture, and deep integration into a customer’s workflow are now the line between a credible pitch and “a prompt with a UI.”

The three problems every agentic AI founder eventually hits

Whichever side of the infra/vertical line you sit on, there are structural challenges in this category that are different in kind, not just degree, from the SaaS era most of us grew up investing in.

1. Token costs are quietly eating the business model. This is the one that catches founders off guard, because it doesn’t show up until you have real usage. AI features are compressing gross margins from the 80-90% SaaS founders are used to down to 50-60%, because every agent action is a metered compute event, not a marginal-cost-zero software interaction. Goldman Sachs projects token consumption could grow roughly 24x, to around 120 quadrillion tokens a month, between 2026 and 2030.

And critically, falling per-token prices have not translated into falling bills, because usage intensity (longer contexts, more agent steps, more autonomous loops) is growing faster than prices are dropping. Several companies have reported blowing through their entire annual token budget in a single quarter. Aaron Levie’s recent post on this, token costs are becoming one of the hottest issues in AI economics captures exactly the dynamic we’re seeing across our portfolio: usage growth outpacing unit economics improvement, and founders discovering that “more usage” isn’t automatically “more margin.”

As tokens take on a significant amount of the cost of any given workflow, then companies will inevitably want to ensure that their dollars go into the most efficient use of tokens for the particular job at hand.

Aaron Levie, Box CEO

The startups we find most compelling right now are the ones treating cost governance and pricing architecture as a product problem from day one, not a finance afterthought to solve post-PMF. Marking up tokens is a race to the bottom; the winners are the ones who design around consumption rather than reselling it.

2. Non-determinism breaks the QA playbook. Traditional software is deterministic: same input, same output, testable, debuggable, repeatable. Agents are not. The same prompt, the same workflow, run twice, can produce two different action sequences, and in a multi-step agentic chain, small probabilistic deviations compound. This isn’t a bug to be fixed; it’s the nature of the technology, and it means founders building in this space need fundamentally different approaches to evaluation, monitoring, guardrails, and rollback than anything in the previous software generation’s toolkit. We’d argue this is the reason the “control plane” sub-category of infrastructure (evals, observability, agent monitoring) has become one of the most actively funded corners of the market: enterprises will not deploy what they cannot predict, audit, or roll back.

3. Security is a different shape of problem entirely. Giving an autonomous system the ability to take actions (call APIs, move money, touch customer data, write and execute code) opens an attack surface that didn’t exist in the chatbot era. Prompt injection, data exfiltration through tool use, identity and permission sprawl across agent fleets, and the simple fact that an agent can be socially engineered the way a human employee can: these are not edge cases, they’re the central design challenge. It’s no surprise that some of the sharpest recent rounds we’ve tracked are security-for-agents plays (Onyx Security’s $35M Series A with Conviction, building on an earlier Cyberstarts seed, is a good regional example). Expect “agent security” to do for this wave what “cloud security” did for the SaaS wave, except compressed into a fraction of the time.

Where we’re spending our time

As an early-stage fund with deep Israeli roots and a thesis around founders building from “the edge” (gaming, consumer, and now agentic systems), we’re drawn to teams that treat these three challenges (cost, non-determinism, security) not as risks to be managed around the edges, but as the actual product surface area. The founders who are going to build enduring companies here are the ones designing for unpredictable, metered, adversarial environments from the first commit, not retrofitting governance once an enterprise customer asks an uncomfortable question in a security review.

We mapped 32 named funds across just ten recent Israeli agentic AI raises, from Team8 and Glilot to Insight Partners, Index Ventures, and Lightspeed, which tells you how fast capital is moving into this space.

See the full breakdown on IsraelVC