Technologie & IA
Fhenix on Ethereum and Arbitrum: understanding CoFHE without confusing it with a rollup
Fhenix presents its confidential-computing infrastructure as live on Ethereum and Arbitrum. CoFHE works with encrypted data; this architecture is not the launch of a new second-layer chain.

Privacy infrastructure on existing networks
As of 22 October 2026, Fhenix's website presents its infrastructure as live on Ethereum and Arbitrum. Its documentation explains that the project enables contracts using encrypted data. The stated interest notably concerns confidentiality of financial information while it is being processed, rather than creating a new network to which all assets would have to be moved.
This presentation must not be confused with launching a new independent blockchain. The current product is called CoFHE and relies on a confidential-computing coprocessor. The network hosting a contract and the infrastructure performing certain calculations are distinct components. Identifying them clarifies where the proposed service sits within an application's operation.
An explicitly documented architectural change
An older official post concerning rollups now includes an update. Fhenix specifies that its initial vision evolved towards CoFHE without requiring a separate new second-layer chain. Reading the post without this notice would attribute to the current product a deployment model that the project itself has revised.
CoFHE should therefore not automatically be described as a system batching transactions to lower fees on a new rollup. Its highlighted contribution is computing with encrypted information. Cost reduction, network availability and application confidentiality answer different technical questions, each requiring its own demonstration.
Computing without public disclosure does not eliminate permissions
Fully homomorphic encryption allows certain operations to be performed on data that remains encrypted during calculation. The documentation distinguishes tools preparing inputs, functions available to contracts and mechanisms for obtaining a readable result. These different stages must therefore not be reduced to the simplified idea that nobody could ever gain access to any information within an application using this technology.
The technical post published on 20 October specifically describes the role of access controls and decryption permissions. Possessing a reference to data is insufficient to be authorised to read it. The confidentiality provided by an application therefore also depends on the way its access rights are assigned to participants and subsequently used, rather than solely on the presence of a cryptographic primitive.
Security work is not an absolute guarantee
In the same post, Fhenix presents work undertaken with Consensys Diligence on confidential applications. This provides information about the security questions examined. It would be excessive to infer that every present or future contract is guaranteed free of vulnerabilities or that each financial use automatically complies with every applicable obligation.
The useful reading examines the current architecture, permissions and documentation of the service actually used. A planned partnership, an available application and infrastructure announced as live do not represent the same stage of development. In this case, verified elements explain CoFHE's role without presenting stablecoin services or partnerships as already completed when their arrangements are not established by the documents consulted.
Distinguishing these stages is necessary to describe an actual technical offering without turning a development intention into a functioning product. It also separates the cryptographic capability itself from the choices made by an application about access to information, the organisation of its service and the rights it grants to particular participants.
