Dream announced on June 18, 2026 that it has raised $260 million in Series C funding at a $3 billion post-money valuation, nearly tripling its February 2025 valuation of $1 billion. Bicycle Capital and Group 11 led the round, with Bain Capital Ventures, Antler, and Tru Arrow Partners participating. The company was founded in 2023 by Shalev Hulio (former NSO Group CEO), former Austrian Chancellor Sebastian Kurz, and Gil Dolev, and now employs about 350 people across Tel Aviv, Abu Dhabi, and Vienna. Dream sells AI-driven cyber defense to governments and critical infrastructure operators and says it has secured close to $300 million in contract value since starting commercial sales in late 2024.
The interesting variable here is not the funding size but the customer list. Dream's pitch is sovereign-grade AI for nations that do not want to route classified telemetry through a US hyperscaler, and the round closed days after Washington forced Anthropic to disable Fable 5 and Mythos 5 access for foreign nationals. That sequencing reads as a market validation: every export-control directive aimed at frontier models becomes a tailwind for vendors selling domestically deployable AI to allied governments. The Hulio-Kurz combination is a deliberate signaling choice — operational depth in offensive cyber paired with diplomatic access at the head-of-state level.
Sovereign AI as a distinct category has gone from buzzword to budget line in roughly twelve months. NVIDIA framed it in early 2025, the EU and Brazil signed a Digital Partnership on June 12, Switzerland's Prem AI confirmed a $100 million Series A on the same day Dream closed, and Korea's MOU with Anthropic last week explicitly cited in-country deployment. Governments are paying for AI infrastructure they can audit, geofence, and turn off — and a generation of cyber firms is repositioning around that demand. Dream sits at the most security-sensitive end of that spectrum.
Takeaway for learners: AI is no longer a single market — it is splitting into a public-internet tier and a sovereign tier with different vendors, different deployment models, and different procurement cycles. If you are evaluating careers or startups in AI, knowing which tier a company sells into matters more than knowing which model it runs on. The sovereign tier rewards regulatory fluency and government access; the public-internet tier rewards distribution and product velocity.