Using threshold secret sharing, we propose a solution tai-lored for forgetful clients (i.e., not required to keep any cryptographicsecret) while accommodating the dynamic nature of multi-cloud deploy-ments. Furthermore, we delegate the computation and distribution ofshares to an intermediate server (proxy), effectively minimizing the clientworkload. We propose two variants of a keyless, space-efficient multi-cloud storage scheme named KAPRE and KAME. Our solution KAPRErequires less communications and computations, while KAME preservesdata confidentiality against a colluding proxy. Our protocols offer robustguarantees for data integrity, and we demonstrate the proxy’s ability toidentify and attribute blame to servers responsible for sending corruptedshares during data reconstruction. We establish a comprehensive secu-rity model and provide proofs of the security properties of our protocols.To complement this theoretical analysis, we present a proof-of-conceptto illustrate the practical implementation of our proposed scheme.