Israel is justified in its concerns over whether scientific, technological and economic value created by artificial intelligence companies in biopharma research and development will remain in the country as those companies develop and grow.
Many such companies start with employees concentrated in Israel. Later, a share of their workforce and senior leadership may be based in the United States. As this trend gains momentum, it eventually means that the company leaves Israel.
But that’s only half the story.
Three key questions linger: What changes as these companies’ AI models and platforms mature? Why do they need clinical, product, leadership and commercial capabilities in the United States? And why can that evolution ultimately create more value for Israel rather than less?
Many Israeli AI companies in biopharma R&D begin with Israeli founders, machine learning specialists, data scientists and engineers. For the platform to mature into a viable product, however, the company must evolve. It needs employees who understand clinical development and translational medicine; the decisions biopharma organizations face; and the uncertainties and risks decision-makers must weigh before advancing an asset.
A company also needs product leaders who understand pharmaceutical R&D workflows and senior executives and employees in business development, marketing and sales with industry networks who can explain the platform’s value.
Many of these capabilities are added in the United States because that is where the relevant talent and market are concentrated.
Consider an example: An Israeli AI company begins with 10 employees in Israel focused on developing, training and validating machine learning models, building data pipelines and engineering the platform. The company later hires 10 employees in the United States involved in clinical leadership, product, business development, marketing, sales and senior management.
Thus, Israel’s share of total employment has fallen from 100% to 50%, but the original Israeli technological organization may remain intact. At the same time, the company has added capabilities that it did not previously have.
Independently, the Israeli technical organization can evolve. As models and platform components become established and validated, work becomes more specialized. In some companies, this may reduce the number of employees required for early development activities.
These developments are too easily interpreted as one phenomenon. Some companies move important functions to the United States, but these situations should not all be described as Israeli AI talent moving to the United States.
For example, such AI platforms must solve problems inside pharmaceutical and biotechnology R&D, from biological and clinical data analysis to biomarker identification, treatment response prediction and clinical trial simulation. Building sophisticated AI models is essential, but it is not sufficient, because a viable company must understand how pharmaceutical R&D works, how development teams make decisions, what evidence they require and how an AI platform fits into existing workflows.
The U.S. organization complements the Israeli base rather than replacing it.
A clinical development executive in Boston may have spent years on the other side of the table at a pharmaceutical company. That executive can understand the decisions and uncertainties that pharmaceutical leaders face; explain the platform’s value in terms relevant to those decisions; and bring credibility and relationships that an Israeli AI team without similar experience may lack. The executive can also feed insights from those discussions back to product, data science and engineering teams.
In fact, these employees can become part of the development process itself. That is why describing the model as “R&D in Israel followed by commercialization in the United States” is too simplistic.
A significant portion of the relevant talent, potential customers and partners is concentrated in the United States. Consequently, an American AI company operates within that ecosystem. On the other hand, an Israeli company can possess competitive technology but lack equivalent access.
In my view, building substantial capabilities in the United States is necessary for many Israeli AI companies in biopharma R&D. For some, it can be the difference between developing impressive AI technology and building a company that survives.
If an Israeli AI company remains entirely in Israel but never develops the scientific, clinical, product and industry capabilities required to become commercially viable, Israel may retain almost all the company geographically while capturing very little value from it.
If important AI, data science, engineering, scientific expertise and intellectual property creation continue in Israel, the capabilities added in the United States can make what is being built in Israel more valuable. Clinical and scientific leaders close to pharmaceutical customers can bring knowledge back into the organization, while product leaders can translate industry needs into requirements that improve the platform.
Commercial success in the United States can feed investment back into Israel. More customers, partnerships and revenue can provide resources for additional model development, data science, engineering and scientific programs in Israel and high-value hiring.
In this case, the U.S. organization complements the Israeli base rather than replacing it.
Israel should be deeply concerned when the opposite happens—and core AI development, intellectual-property creation, scientific leadership and other high-value capabilities progressively disappear. But that is a different phenomenon.
The objective should be to reduce the ecosystem access disadvantage faced by Israeli founders while encouraging companies to continue deepening AI, data science, engineering and scientific capabilities in which the Jewish state has genuine strengths.
A growing U.S. share of employees can make an Israeli AI company appear less Israeli, but for many Israeli AI companies in biopharma R&D, it means that the company is evolving. Its models and platform are maturing. The capabilities it needs are changing. It is adding scientific, clinical, product, leadership and commercial expertise in the United States because that is where much of the relevant talent, market and biopharma ecosystem are located.
The real measure should be whether the company acquires the capabilities it needs to survive and compete while continuing to create scientific, technological and economic value in Israel.
If it does, what may appear as a loss could be the process through which the company becomes successful enough for Israel to benefit from its success.
And that is the other half of the story.
