
VC AI Survey
Elephant CEO on Israel’s AI future: Excelling in defense, but missing the LLM wave
For CTech’s VC AI Survey, CEO Itay Ben Ari spoke about the impact of artificial intelligence and how Startup Nation can catch up with new trends.
“In Israel, I see artificial intelligence increasingly shaping the defense, cyber, and homeland security sectors—particularly given the technological warfare of the past two years and our proven edge over adversaries,” said Itay Ben Ari, CEO of Elephant Brokerage. “As governments across Europe ramp up defense budgets, they’re likely to prioritize technologies that have been battle-tested. This positions Israel’s new warfare AI-driven systems and capabilities to become a major trend in the coming years.”
Ben Ari joined CTech for its VC AI Survey to share insights on how Israel can contribute to the growing need for AI tech across the world. Specifically, he outlines where the country has succeeded - and where it needs to play catch-up.
“Currently, there are no companies in Israel developing Generative AI or large language models,” he added. “But it seems that in this domain, the train has already left the station, and we’ve missed it - especially when considering the vast data and energy resources required to train a new model.”
You can learn more in the interview below.
Fund ID
Name and Title: Itay Ben Ari, CEO of Elephant Brokerage
Fund Name: Elephant Secondary Platform
Founding Year: 2015
Investment Stage: A platform that connects sellers and buyers in late-stage private tech companies
Investment Sectors: Secondaries
On a scale of 1 to 10, how has AI impacted your fund’s operations over the past year - specifically in terms of the day-to-day work of the fund's partners and team members?
8 - Over the past year and a half, the AI sector has attracted the most interest from our investors. Companies operating along the entire value chain (not just those developing LLM and Generative AI models) hold a significant place on their target lists for potential investments or secondary share purchases.
A significant portion of our resources is directed toward identifying early investors and stakeholders looking for liquidity in these companies and securing allocations in primary rounds of leading companies, where investor demand exceeds available supply.
Have you already had any significant exits from AI companies? If so, what were the key characteristics of those companies?
The majority of investments by Elephant Group’s clients were made over the past three years. Given that some of these companies have only been operating for a relatively short period (despite their high valuations and the attention surrounding them), there haven’t been many exits yet—aside from early investors who chose to sell shares through secondary transactions.
However, the example of CoreWeave, which went public this year and has seen its valuation more than triple since its IPO in March 2025, may be an early indicator of a broader trend emerging in the sector. The company operates in the field of cloud infrastructure, providing computing resources specifically for artificial intelligence and machine learning projects—a segment of the value chain that many investors are actively targeting within the AI space.
Is identifying promising AI startups different from evaluating companies in your more traditional investment domains? If so, how does that difference manifest?
From what we’re seeing with our institutional clients, the core process remains the same. However, with some of them, we’re observing adjustments to specific parameters in their models in order to better capture the state of the market, the pace of growth, and the initial inflection point of its development.
What specific financial performance indicators (KPIs) do you examine when assessing a potential AI company? Are there any AI-specific metrics you consider particularly important?
It varies and depends on where the company operates within the AI value chain. For example, looking at companies who operate in the LLM and Generative AI space, KPI relevant for the space:
The key performance indicators (KPIs) for AI-driven companies span several critical categories. Revenue and growth metrics include annual revenue, year-over-year (YoY) growth, annual recurring revenue (ARR), and the extent of enterprise adoption. User metrics track engagement and retention through monthly and daily active users (MAUs and DAUs) across platforms. Infrastructure spend reflects GPU utilization, cloud expenses, and data center expansion. Strategic partnerships are measured through collaborations with hyperscalers, enterprise clients, and government contracts.
Finally, model performance is evaluated based on benchmark scores such as MMLU and BBH, as well as latency, context window size, and overall accuracy.
In the case of AI hyperscaler companies in the cloud infrastructure space, I would add:
- Revenue Backlog (TCV) – Total Contract Value, represents expected revenue from signed contracts and is used to forecast cash flow and revenue recognition over time
- Adjusted EBITDA Income/Margin – due to high expenditures needed for Infrastructure Scaling, this is a good indicator for operational efficiency.
- Infrastructure Scaling – Companies in the SaaS computing space rely heavily on expanding their compute capacity. This growth is critical not only for fulfilling contractual obligations but also for converting Total Contract Value (TCV) into Annual Recurring Revenue (ARR) and ultimately recognized revenue.
- Debt-to-Equity Ratio – This metric reflects a leveraged capital structure, which is common among infrastructure-intensive companies with high capital expenditures. A high ratio often indicates aggressive scaling and serves as a key indicator of financial risk exposure but also for capital equity efficiency.
How do you approach the valuation of early-stage AI startups, which often lack significant revenues but possess strong technological potential?
We operate very rarely in early-stage companies that do not yet generate revenues at a level that allows our investors to assess the company’s financial matrix.
What financial risks do you associate with investing in AI companies, beyond the usual technological risks?
All the risks mentioned above are certainly important and valid. I would add two more. The first relates to the rapid pace of development we’re seeing in the market—every day, a new company emerges offering something newer and better than those before it. The second concerns the growing energy consumption required to train each new model that enters the market and the accelerating rate at which this demand is increasing.
Do you focus on particular subdomains within AI?
We are agnostic to any specific sub-sector and provide our services according to the needs of our clients.
How do you view AI’s impact on traditional industries? Are there specific AI technologies you believe will be especially transformative in certain sectors?
Artificial intelligence is poised to reshape nearly every aspect of our lives. As organizations embed AI more deeply into their operational environments, tailored to their workflows, data systems, and operations, its impact will accelerate—driving efficiency and adaptability across industries. Tasks that currently require significant human effort and cost will increasingly be handled by a fleet of autonomous, self learning and adaptive agents, delivering faster and more cost-effective results. For businesses and shareholders, this shift promises greater productivity and enhanced profit margins. Yet, it also raises serious challenges for workers in roles that may become obsolete, underscoring the need for thoughtful transition strategies.
What specific AI trends in Israel do you see as having strong exit potential in the next five years? Are there niches where you believe Israeli startups particularly excel?
In Israel, I see artificial intelligence increasingly shaping the defense, cyber, and homeland security sectors—particularly given the technological warfare of the past two years and our proven edge over adversaries. As governments across Europe ramp up defense budgets, they’re likely to prioritize technologies that have been battle-tested. This positions Israel’s new warfare AI-driven systems and capabilities to become a major trend in the coming years.
Are there gaps or missing segments in the Israeli AI landscape that you’ve identified? What types of AI founders are you especially looking to back right now in Israel?
Currently, there are no companies in Israel developing Generative AI or large language models (LLMs). But it seems that in this domain, the train has already left the station, and we’ve missed it- especially when considering the vast data and energy resources required to train a new model.