Exclusive: NIF backs defence startup Terasi in €11m raise
THE SO WHAT
Defence capital is moving earlier in the stack—€11M at this stage suggests governments want optionality on novel capabilities before they’re proven. If you’re building dual-use tech in Europe, assume procurement and export-control conversations now start at seed, not Series B.
READ THE SOURCE
MORE FROM THE WIRE
Startups & VentureExhibit tables added: One last chance to showcase your startup at TechCrunch Disrupt 2026
Late-stage exhibitor reopenings at Disrupt are a reminder that attention — not just capital — is oversubscribed. If you’re early-stage and actually need customers or hires from SF in October, this is a tactical slot to test your pitch and messaging density in front of 10,000+ industry participants.
Startups & VentureAndreessen Horowitz Launches AI School as a College Alternative
A16z standing up an AI-focused college alternative is a bet that the scarcest resource in the stack is founder-quality talent, not capital or compute. If you hire early-career builders, expect more candidates coming from non-traditional programs with strong networks but no formal credentials—adjust filters accordingly.
Startups & Venturea16z invests $35M in Horowitz Andreessen Academy, an independent, unaccredited two-year alternative to college; Anthropic, Meta, and others will provide compute
A16z putting $35M into a two-year, unaccredited AI academy with compute from Anthropic, Meta, and others formalizes a parallel talent pipeline outside universities. If you’re competing for junior AI talent, you’re now competing with VC-backed training programs that bundle education, network, and cloud credits.
Startups & VentureNscale’s IPO will test Wall Street’s appetite for concentrated AI bets once again
A data center developer whose revenue is tightly coupled to Microsoft and Anthropic is a public-market test of how much counterparty and platform risk investors are willing to underwrite for AI infra growth. If this prices well, expect more upstream providers—power, land, cooling—to lean into similar concentrated AI exposure rather than diversify.