Nillion

Nillion's team has delivered a working mainnet and the NIL token on schedule in March 2025, which clears the basic execution bar, but the project still reads as developer-stage rather than product-stage. The roadmap has favored protocol primitives over killer applications, leaving the team's commercialization ability unproven even as its cryptographic engineering looks credible. Against Kaito, which ships consumer-facing products and pivots under pressure, Nillion's cadence is steadier but less demonstrably demand-driven. Lookout would re-rate on a flagship integration where blind compute is load-bearing for a real, paying workload rather than a testnet showcase.

Nillion's 'blind computing' — MPC-based computation over data that never gets decrypted — is technically legitimate and arguably more practical than full FHE for many use cases. The problem is the field: it sits in a crowded FHE/MPC privacy-infrastructure race against Zama, Fhenix, Inco, and others all chasing the same encrypted-compute thesis, so being competent is not the same as being differentiated. Versus Zama's FHE-first approach, Nillion bets MPC delivers usable performance sooner, a defensible but contested wager. Lookout would upgrade on independent benchmarks showing Nillion's blind compute meaningfully out-performs FHE rivals on latency or cost for a concrete workload.

The NIL token's value depends on metered demand for private computation — a market that today barely exists at commercial scale, so the economics are forward-priced on a thesis rather than current usage. With privacy infrastructure still pre-revenue across the sector, the token captures speculative network value more than fees, and unlock schedules add supply pressure against thin organic demand. Compared to a name with live consumer revenue streams, NIL's accrual story is notably thinner and more dependent on a future inflection. Lookout would revisit if on-chain compute fees or a clear staking-secures-real-work mechanism started showing non-trivial, growing volume.

Adoption remains the soft spot: Nillion has builders and grants but little evidence of production applications generating sustained, paid blind-compute demand. Privacy is a famously hard sell — developers say they want it, then ship without it — and Nillion has not yet shown the killer use case that converts interest into recurring usage. Against Kaito's observable daily engagement, Nillion's traction is aspirational. Lookout would shift its view on a marquee deployment — an AI data marketplace, a private-inference product, or a regulated-data use case — processing real volume through the network.

The ~$50M Series A with Hack VC is a serious war chest by AI-crypto standards — roughly five times Kaito's Series A — giving Nillion years of runway to find product-market fit without forced token-side pressure. Hack VC's deep infrastructure thesis lends credibility and a network of integration partners in the privacy and AI-data space. That capitalization stands well above thin-syndicate peers like Entangle or Atlas Navi and is the clearest single strength on the scorecard. Lookout sees no concern here; the question is conversion of capital into demand, not access to capital.

Privacy-preserving compute for AI data is a genuinely durable narrative — every serious AI-on-chain discussion eventually hits the confidentiality wall Nillion targets. But the narrative is also diffuse and shared by a dozen FHE/MPC projects, so Nillion does not own the story the way Kaito owns InfoFi. The thesis is real but the category is loud, leaving Nillion as one voice among many rather than the reference name. Lookout would warm to it if Nillion became the default 'private compute layer' that AI-agent and data projects cite by name, rather than a generic privacy option.

Nillion's principal risk is demand-timing: the technology may simply be early, and a well-funded protocol can burn runway waiting for a market that arrives years late or routes around it. Competitive risk compounds this — if FHE matures faster than expected, the MPC bet could age poorly, and the crowded field means a rival could capture the privacy mindshare first. Relative to platform-dependent names like Kaito, Nillion owns its network and carries less third-party switch-off risk, a genuine offset. Lookout would lower its risk read on early evidence of real paid demand that anchors the token to usage instead of speculation.