The Mossad’s New Portfolio: SoftBank’s Bet on AI Security as the Ultimate Narrative Capital

When SoftBank announced that former Mossad chief Yossi Cohen would join as a strategic advisor for AI investments, the market didn’t blink. It should have. This isn’t a routine hire—it’s a signal that the world’s largest tech investment vehicle is reframing the AI race from a battle of algorithms to a contest of trust and geopolitical leverage. Where digital pixels breathe with human soul, the line between intelligence and intelligence agency is blurring. SoftBank is no stranger to bold pivots. After the WeWork debacle and the Vision Fund’s contraction, Masayoshi Son doubled down on AI, betting the firm’s future on an AGI-driven infrastructure play anchored by Arm Holdings. The chip architecture powerhouse gives SoftBank a unique vantage point over the entire AI stack—from silicon to sovereignty. Yet the missing piece was always the human factor: how to assess the trustworthiness of AI startups, how to navigate the murky waters of export controls, and how to secure the narrative that SoftBank’s AI capital is safe for sovereign funds. Enter Cohen, a man whose career was built on mapping the unseen currents of narrative capital. Cohen’s appointment is more than a resume addition. It’s a recognition that the highest-stakes AI investments are no longer purely technical but deeply political. The core insight is that SoftBank is building a new competitive moat: a fusion of capital, chip control, and intelligence-grade security assessment. This is not about hiring a compliance officer; it’s about embedding a former spy chief into the M&A process. The logic is subtle but potent. In DeFi, I learned that oracle feed latency is the Achilles’ heel—security is not a feature, it’s a foundation. Similarly, in AI, the most dangerous vulnerability is not a model hallucination but a supply chain backdoor that a nation-state could exploit. Cohen’s network can potentially vet startups for such risks, especially those emerging from Israel’s elite intelligence units like Unit 8200. Based on my own experience auditing the Gnosis Safe multisig in 2017, I know that the most critical vulnerabilities often hide in the social layer, not the code. Cohen’s role is to audit the social layer of SoftBank’s AI portfolio. But here lies the contrarian angle: this move could backfire spectacularly. The same network that opens doors to Israeli defense tech may slam them shut in China, where SoftBank still has legacy investments. European regulators, already wary of AI’s surveillance potential, may view the appointment as a red flag. And the AI research community—a community that values independence and ethical boundaries—may recoil from associating with a former intelligence chief. SoftBank risks becoming a target of the very trust deficit it seeks to mitigate. The narrative capital Cohen brings is a double-edged sword: it can unlock deals, but it can also poison the well of public trust. In the DeFi world, we saw how a single vulnerability in a centralized oracle could shake an entire ecosystem. Here, the vulnerability is perception. SoftBank is betting that the market will value security over autonomy, but history suggests that crypto’s ethos—and by extension, AI’s open-source spirit—fights back against centralized authority. Where digital pixels breathe with human soul, the next AI narrative is not about who builds the best model, but who can navigate the minefield of ethics, regulation, and espionage. SoftBank is placing a massive bet on a new form of capital: one that is as much about intelligence as it is about intelligence agencies. The question is whether the market will buy that story—or whether the unseen currents of narrative capital will drag the firm into a contradiction too deep to resolve. In the end, the most important audit is not of code, but of the human heart. SoftBank’s heart now beats with a pulse from the Mossad. We are mapping the unseen currents, and they are powerful—but they are also perilous.