Adi Shamir: Building the Foundations of Modern Public-Key Cryptography

Adi Shamir is an Israeli cryptographer and researcher whose work helped establish modern information security. After receiving his mathematics degree from Tel Aviv University and completing graduate degrees in computer science at the Weizmann Institute, he conducted research at the University of Warwick and the Massachusetts Institute of Technology.

Establishing RSA and Enterprise Cryptography

In 1982, Shamir joined Rivest and Adleman to establish RSA Security, a company dedicated to developing encryption and decryption standards. Headquartered in Burlington, Massachusetts, the firm created widely used commercial security products, including the SecurID authentication token and the BSAFE cryptography libraries. The corporate footprint of RSA also expanded globally, establishing regional offices in Bracknell in the United Kingdom and in Singapore. The enterprise drew substantial commercial interest across subsequent decades: EMC Corporation acquired RSA in 2006 for 2.1 billion dollars, followed by its integration into Dell Technologies in 2016, and an eventual 2.1 billion dollar acquisition in 2020 by an investment group including Symphony Technology Group, the Ontario Teachers’ Pension Plan Board, and AlpInvest Partners.

Inventing Protocols and Breaking Ciphers

While maintaining his academic appointment at the Weizmann Institute starting in 1980 and later joining the École Normale Supérieure in Paris, Shamir continued creating essential mathematical tools for cryptography. He introduced the Shamir secret sharing scheme and, in 1984, proposed identity-based cryptography, allowing public keys to derive directly from personal identifiers like email addresses. In 1986, working alongside Amos Fiat, he devised the Fiat-Shamir heuristic, an approach that transforms interactive identification protocols into digital signatures. His collaborative contributions also generated the Feige-Fiat-Shamir identification scheme and, in 2001, the concept of ring signatures, which permit members of a group to sign records without disclosing individual identities.

Alongside constructing new defenses, Shamir repeatedly evaluated existing systems. In the late 1980s, alongside Eli Biham, he developed differential cryptanalysis, a comprehensive method for analyzing block ciphers. He also successfully compromised the Merkle-Hellman knapsack cryptosystem. In 2000, Shamir, Alex Biryukov, and David Wagner unveiled a real-time attack on A5/1, the stream cipher safeguarding GSM mobile communications. A year later, with Scott Fluhrer and Itsik Mantin, he demonstrated the Fluhrer-Mantin-Shamir attack against the RC4 stream cipher, exposing a critical vulnerability in the Wired Equivalent Privacy protocol used by early Wi-Fi networks.

Investigating Physical Implementations and Side Channels

Shamir expanded cryptanalytic investigations beyond mathematical formulations by targeting the physical environments in which algorithms execute. He created hardware concepts such as the TWINKLE and TWIRL factoring devices, developed visual cryptography, and investigated side-channel vulnerabilities, including cache-based attacks. In 2014, working with Daniel Genkin and Eran Tromer, he proved that cryptographic keys could be collected through acoustic analysis. By monitoring the acoustic signals generated by a laptop during decryption routines, the team extracted full 4096-bit RSA keys, illustrating how operational emissions could undermine secure software.

Computational Theory and Artificial Intelligence

Shamir has also advanced foundational computer science and emerging computational systems. His theoretical research yielded the first linear-time algorithm for 2-satisfiability and established the mathematical equivalence of the complexity classes PSPACE and IP. Since the late 2010s, he turned his focus toward machine learning robustness, examining adversarial examples and formulating the Dimpled Manifold Model to explain how neural network boundaries change during training. His team also formulated cryptanalytic approaches to reconstruct the internal parameters of neural networks. His broad contributions earned him the ACM Turing Award in 2002 alongside Rivest and Adleman, followed by major honors including the Israel Prize, the Japan Prize, the Wolf Prize in Mathematics, and the Levchin Prize.