KEY INSIGHTS
- The Google MPC security system uses distributed key control to ensure financial firms can transfer digital assets safely from any security threats.
- Google is currently piloting the MPC security system with large banking firms before its planned early 2027 launch date.
- The Google MPC security system supports post-quantum cryptography using transaction policies and upgrades.
Google is creating the Google Multi-Party Compute security system for protecting the digital assets of banks and financial institutions and allowing them to deal with cryptocurrencies and blockchain transactions. Google will make the product available on the market in early 2027. Alphabet, Google’s parent company, has also started using the technology to secure its digital assets on Ethereum and Solana.
The service addresses the challenge of protecting private keys, which authorize digital-asset transfers. Google has not identified its testing partners, disclosed which assets Alphabet is securing through the system, or announced pricing.
How Google Multi-Party Compute Protects Digital Assets
Google Multi-Party Compute uses multi-party computation (MPC), a cryptographic approach that distributes control of a private key among separate participants. Instead of keeping the entire private key in one place, the system is designed such that various components cooperate in authorizing transactions.
Source: Watcher.Guru
This design is meant to minimize the dangers of a breach of a particular system or credential. The attacker who breaches one system will not necessarily have the entire private key that is required for authorizing a transaction.
Google’s service integrates distributed signing with transaction policies and key recovery. While its existing Google Cloud Key Management Service and Cloud HSM offerings have the capability to manage cryptographic keys, the new product has transaction authorization and recovery for digital assets.
In 2023, Google published technical guidance explaining how organizations could build MPC-based systems using its Confidential Space infrastructure. The proposed service brings related capabilities together in a dedicated product.
Transaction Policies Add Institutional Controls
Financial institutions using the service could distribute signing credentials among Google Cloud, their own hardware, and other cloud providers. Google says its architecture supports different approval thresholds, including arrangements requiring two of three participants or five of seven participants to authorize transactions.
Before signing begins, a policy engine operating within Google’s Confidential Space environment checks transactions against institutional rules. These rules could include transfer limits, restrictions on destination wallet addresses, and approval requirements involving multiple departments.
Google also says customers retain control of their key material and that the company cannot access it in unencrypted form, even when key shares are stored on its infrastructure.
Nevertheless, credential distribution is not a remedy for all risks that exist. Security will depend on how the system will be set up by the institutions, how well their devices and administrators will be secured, and how they control any modifications to the requirements of authorization.
Private-Key Compromises Persist as Security Risk
The proposed service is being offered in a time of ongoing losses associated with compromised cryptocurrencies.
In March 2026, the stablecoin protocol Resolv was hacked due to a compromised private key having access to an authorized account and problems with the minting process. The incident allowed an attacker to extract approximately $25 million.
In June, attackers compromised private keys belonging to a Humanity Protocol foundation member and drained more than $32 million from at least 17 wallets. Furthermore, the hacking incident coincided with a drop of more than 80 percent in the value of the tokens in the project.
Google’s planned entry places it alongside specialized digital-asset security providers, including Fireblocks, Fordefi, and BitGo.
Security Concerns Related to AI and Quantum Computing
Google also noted cybersecurity concerns related to artificial intelligence and quantum computing. These concerns are distinct from usual attacks on credentials, and a distributed signing approach does not automatically solve all weaknesses of cryptographic systems.
According to Chainalysis, AI-tool vendor related scams yielded average proceeds of $3.2 million per incident in 2025 compared to only $719,000 for other scams. The figures show a difference between the two groups but do not establish that AI directly caused the higher proceeds.
Ethereum researcher Justin Drake has warned that future AI advances could potentially identify mathematical shortcuts affecting the digital signatures used by Bitcoin and Ethereum. No AI system has publicly demonstrated the ability to break those underlying cryptographic signatures.
Google says its architecture is designed to accommodate future post-quantum cryptography upgrades. The service is expected to support ECDSA and Ed25519 signing algorithms at general availability, although the company has not announced a timetable for introducing post-quantum signing capabilities.
