Google is previewing Google Multi-Party Compute, a crypto custody system that splits private signing keys across separate environments, at its first digital assets symposium in New York. The product is currently in private preview with several financial institutions and is scheduled for commercial availability in early 2027, though pricing has not been disclosed. Alphabet has reportedly begun using the technology to secure its own internal assets.
The system utilizes multi-party computation to divide keys into fragments held by different parties or machines, ensuring the full key is never assembled in one location. Institutions can configure thresholds such as two-of-three or five-of-seven signatures required for transaction approval. A policy engine within Google’s Confidential Space validates transfers against limits and destination restrictions before signing occurs. Google states it does not have access to key shares in unencrypted form.
This development addresses persistent security vulnerabilities in the cryptocurrency sector. Chainalysis attributed 43.8% of cryptocurrency stolen in 2024 to compromised private keys, while DeFiLlama data indicates roughly 40% of losses through June came from similar breaches. Google also positions the tool against AI-enabled fraud, citing Chainalysis findings that scams linked to AI vendors averaged $3.2 million per operation in 2025 compared to $719,000 for other scams. Additionally, the architecture is designed to potentially adopt post-quantum signatures later, responding to long-term risks identified by Google researchers regarding quantum computing's ability to derive private keys.
Google’s entry into institutional crypto custody signals a strategic shift toward leveraging cloud infrastructure neutrality to compete with specialized providers like Fireblocks and BitGo. By offering a solution where banks can hold key shares on their own hardware while utilizing Google Cloud for others, the company aims to alleviate concerns about vendor lock-in and single points of failure. This approach directly targets the operational risk highlighted by recent high-profile hacks, positioning security architecture rather than just storage convenience as the primary value proposition for traditional finance institutions.
However, the timeline for commercial availability in early 2027 suggests a cautious rollout, allowing time for rigorous testing by pilot partners. The emphasis on future-proofing against quantum threats and AI-driven fraud reflects an attempt to address emerging regulatory and technical anxieties before they become immediate barriers to adoption. While Google claims no access to unencrypted keys, institutional trust will depend heavily on independent verification of these cryptographic guarantees and the stability of the policy enforcement mechanisms during the transition from private preview to general market release.
