In order to solve the “impossible triangle” problem of the blockchain, Silvio Micali, a 64-year-old Turing Award winner and professor of the Massachusetts Institute of Technology’s Computer Science and Artificial Intelligence Laboratory (MIT CSAIL), proposed the Algorand project. According to its white paper, Algorand is a “true decentralized blockchain”, where they claim that their decentralized design approach is different from other existing blockchain solutions.
Algorand uses a variation of Proof of Stake(PoS) for their consensus protocol where transactions on the Algorand network are recorded on blocks generated by an independent committee that is randomly selected from a collection of all users(or “miners”). In the committee selection process, Algorand’s unique approach is the encryption lottery, which means that the user is the only one who knows whether he or she is a member of the committee. Any potential malicious participant would not know who the other committee members are, so they cannot bribe or launch a denial of service attack. Algorand replaces the block producer every time, which means there is no need for a fixed committee — which in turn makes it even more difficult for a malicious user to manipulate the transaction history. The emergence of Algorand has attracted widespread attention in the industry, and because of its innovative technology and support from an outstanding team, the project is considered a new benchmark in the blockchain industry.
The emergence of Algorand was a source of debate in the academic world — many scholars and experts have carried out in-depth and detailed research on the project, including Professor Wang. After a period of rigorous argumentation, Professor Wang came to a conclusion different from the white paper of Algorand which he explains in his academic paper.
Algorand’s fatal flaws
In the paper from Professor Wang, he points out that the consensus protocol of Algorand makes few fatal assumptions, and their hypothetical precondition cannot assure the safety of the whole system.
Flaw 1: In the licensed Algorand environment, it is assumed that at least 2/3 of the users are honest, and in the Algorand environment with no entry threshold, it is assumed that at least 2/3 of the currency units are honest. Under these assumptions, the probability of the bifurcation of Algorand block chain is at most 1/109.
Flaw 2: Algorand protocol assumes that most users (or nodes with most currency units) are honest. In particular, it assumes that all honest users will not reveal their ephemeral keys and the keys will be destroyed after they have reached their purpose.
Elaborating on the above two misunderstandings, Professor Wang Yongge explains a detailed argument in the paper. He believes that in the Algorand protocol, if the attacker controls 1/3 of the currency, then they can effectively maliciously fork the blockchain. Meanwhile the assumption that the honest node will not leak its own key is not established. Being driven by malicious financial interest, any node underneath may have the motivation to leak a temporary key which will harm the entire ecosystem.
The Algorand team official has not yet responded to the matter. According to the latest news, Algorand’s main network technology is ready, and the main-net will be rolled out slowly, beginning with small trusted network of users. Professor Wang Yongge’s question about Algorand will be verified by facts once the protocol is fully public, and I believe that in time the “truly decentralized” network will be tested.
Kinsa Durst is the Marketing Manager at SperaX, the first decentralized, post-quantum secure protocol on open-source hardware nodes.