top of page

Encriptación Homomórfica por Cambio Aleatorio de Variables

Desbloqueando el Procesamiento Seguro de Datos

​​

En el mundo actual impulsado por datos, la necesidad de equilibrar seguridad, privacidad y eficiencia se ha vuelto fundamental. Los métodos de encriptación tradicionales protegen los datos en reposo y durante su transmisión (Encriptación de Extremo a Extremo), pero requieren desencriptación para su procesamiento, lo que introduce vulnerabilidades. La Encriptación Homomórfica (HE, por sus siglas en inglés) cierra esta brecha al permitir realizar cálculos directamente sobre datos encriptados, generando resultados que permanecen encriptados hasta que son descifrados de manera segura por usuarios autorizados.

 

Nuestro esquema patentado de Encriptación Totalmente Homomórfica (FHE, por sus siglas en inglés) revoluciona este campo al ofrecer mejoras significativas en eficiencia y aplicabilidad. Basado en un novedoso método de "cambio aleatorio de variables", este esquema supera desafíos comunes como el ruido, los retrasos en el tiempo y el alto consumo de energía, haciéndolo adaptable a aplicaciones reales en diversas industrias.

​​

​

Por Qué la Encriptación Homomórfica es Importante

​

  • Mayor Privacidad y Seguridad
    Con HE, los datos sensibles nunca necesitan ser desencriptados para ser procesados, lo que garantiza privacidad y seguridad completas, incluso cuando son manejados por sistemas de terceros o entornos en la nube. Por ejemplo:

    • Los bancos pueden procesar solicitudes de préstamos sin exponer detalles financieros.

    • Los modelos de IA pueden entrenarse en conjuntos de datos encriptados sin revelar información sensible.

  • Cumplimiento Regulatorio
    HE se alinea con estrictas normativas de privacidad de datos como el GDPR y el HIPAA, al garantizar la confidencialidad de la información personal y sensible durante todo su ciclo de vida.

  • Computación Privada en la Nube
    A medida que crece la adopción de la nube, las empresas son cautelosas al confiar datos sensibles a proveedores externos. HE permite a las organizaciones aprovechar la escalabilidad y rentabilidad de la computación en la nube sin comprometer la seguridad.

  • Aplicaciones en Diversas Industrias

    • Finanzas: Evaluar transacciones financieras encriptadas de manera segura, garantizando que no se exponga información sensible.

    • Salud: Realizar cálculos en registros médicos encriptados para diagnósticos o investigaciones, protegiendo la confidencialidad del paciente.

    • Inteligencia Artificial y Aprendizaje Automático: Entrenar y desplegar modelos utilizando datos encriptados, evitando violaciones o usos indebidos de información.

    • Gobierno y Defensa: Procesar datos clasificados o sensibles de manera segura sin riesgo de exposición.

 

​

Los Beneficios de Nuestro Esquema FHE

​

  • Solución a los Desafíos de Ruido y Eficiencia
    Los métodos tradicionales de HE suelen enfrentar problemas relacionados con el ruido computacional y la ineficiencia. Nuestro método elimina estos obstáculos, permitiendo cálculos de alta precisión sin compromisos significativos en velocidad o consumo de energía.

  • Flexibilidad y Escalabilidad
    Nuestra solución de HE puede adaptarse a diversos requisitos de seguridad, permitiendo a las empresas integrarla sin problemas en sus flujos de trabajo existentes y escalarla según sea necesario.

 

​

Impacto y Potencial de Mercado

 

El crecimiento exponencial de los datos, junto con el aumento de las amenazas cibernéticas, ha generado una demanda urgente de soluciones robustas de encriptación. Al habilitar el procesamiento seguro de datos sin desencriptarlos, la Encriptación Homomórfica está lista para redefinir la seguridad de los datos y desbloquear nuevas posibilidades en la era de la IA y la computación en la nube.

​

  • El mercado global de encriptación homomórfica está proyectado a crecer rápidamente, impulsado por su papel crítico en la privacidad y seguridad de los datos.

  • La adopción temprana proporciona a las empresas una ventaja competitiva, no solo cumpliendo normativas y reduciendo riesgos, sino también abriendo nuevas oportunidades en innovación basada en datos seguros.

​

​​​​

  1. Propuesta

  2. Proof-of-Concept

  3. Conferencia

  4. Enlaces Adicionales​

Lattice-based cryptography: The tricky math of dots
08:39

Lattice-based cryptography: The tricky math of dots

Lattices are seemingly simple patterns of dots. But they are the basis for some seriously hard math problems. Created by Kelsey Houston-Edwards (www.kelseyhoustonedwards.com) Support the channel on Patreon: https://www.patreon.com/ChalkTalkMath Sponsored by Wire (www.wire.com) ________ Post-Quantum Cryptography: https://youtu.be/_C5dkUiiQnw Learning with Errors: Coming January 5, 2023 ________ Timestamps 0:00 - Post-quantum cryptography introduction 0:58 - Basis vectors 1:55 - Multiple bases for same lattice 2:50 - Shortest vector problem 4:32 - Higher dimensional lattices 4:59 - Lattice problems 5:55 - GGH encryption scheme 8:00 - Other lattice-based schemes ________ What more information about lattice-based cryptography? Check out the many resources as https://thelatticeclub.com/ Lattice-based cryptography (blog, part 1): https://writing.chelseakomlo.com/gentle-introduction-lattice-crypto/ Lattice-based cryptography (blog, part 2): https://writing.chelseakomlo.com/a-gentle-introduction-to-lattices-and-lattice-based-key-exchanges-part-2/ GGH-Encryption Scheme (Wiki): https://en.wikipedia.org/wiki/GGH_encryption_scheme Security of GGH Encryption Scheme (research paper): https://link.springer.com/chapter/10.1007/3-540-48405-1_18
Intro to Homomorphic Encryption
52:55

Intro to Homomorphic Encryption

The Private AI Bootcamp offered by Microsoft Research (MSR) focused on tutorials of building privacy-preserving machine learning services and applications with homomorphic encryption (HE). Around 30 PhD students were invited to gather at the Microsoft Research Lab in Redmond on Dec 2nd – 4th, 2019. The program contents were specifically designed for training. Participants mastered the use of HE, the Microsoft SEAL library, and the methodology behind building privacy-preserving machine learning solutions. As a project as well as a competition, students worked in teams, designed and pitched a novel technology built upon what they had learned during the bootcamp, and received feedback and scores from experts. There was a secret prize for the winning team. There were also social events where all participants had a chance to meet and network with other PhD students and experts from MSR. Wei Dai is a senior research SDE in the Cryptography and Privacy Research group at Microsoft Research. His research interests are cryptography, privacy, and high-performance computing, with a focus on homomorphic encryption and hardware acceleration. He received a PhD degree from Worcester Polytechnic Institute in 2019 and a BS degree from Fudan University in 2012. See more from Microsoft Research's Private AI Bootcamp: https://www.microsoft.com/en-us/research/event/private-ai-bootcamp/
A Decade (or so) of fully homomorphic encryption, by Craig Gentry
01:19:11

A Decade (or so) of fully homomorphic encryption, by Craig Gentry

Invited talk by Craig Gentry, presented at Eurocrypt 2021 Abstract: This talk is about the past, present and future of fully homomorphic encryption. I will try to convince you that, over the last decade (or so), FHE has made an amazing transformation from proof of concept to a tool that is actually practical and usable. I will review a few homomorphic encryptions schemes (including non-fully homomorphic ones) to demonstrate commonalities in how we start from a homomorphism, and use masking techniques to defeat common attacks (especially linear algebra) on the homomorphism, to obtain semantically secure homomorphic encryption. And, finally, I will try to delineate the search space for fundamentally new FHE schemes that fall outside of the current framework.
Homomorphic Encryption with Microsoft SEAL
52:58

Homomorphic Encryption with Microsoft SEAL

Since the invention of its first scheme in 2009, homomorphic encryption has been making it possible to perform computations on encrypted data, providing an opportunity to offer greater security assurances to customers using and storing their personal information in the cloud. But implementations of the technology haven’t always been accessible to those creating services people have come to rely on, requiring developers have an expertise in cryptography. In 2015, Microsoft Research released Microsoft Simple Encrypted Arithmetic Library, or Microsoft SEAL, an easy-to-use homomorphic encryption library written in C++ with no external dependencies. Available publicly on Github the library offers detailed examples on how to properly and securely use the technology. In this webinar led by researcher Kim Laine of the Cryptography and Privacy Research group at Microsoft, you’ll learn how SEAL can help software engineers develop data storage and computation services that customers can feel safe using because their personal information is never exposed. Together, you'll explore: ■ The three stages of data privacy; ■ The motivation behind and the evolution of Microsoft SEAL; ■ The core operations of Microsoft SEAL, including rotations of encrypted vectors and serialization; ■ The feasibility of building real-world applications. 𝗥𝗲𝘀𝗼𝘂𝗿𝗰𝗲 𝗹𝗶𝘀𝘁: ■ Microsoft SEAL (project page) - https://www.microsoft.com/en-us/research/project/microsoft-seal ■ Installing Microsoft SEAL on Linux & macOS (walk-through) - https://www.microsoft.com/en-us/research/video/installing-microsoft-seal-on-linux-macos ■ Installing Microsoft SEAL on Windows (walk-through) - https://www.microsoft.com/en-us/research/video/installing-microsoft-seal-on-windows ■ Cryptography and Privacy Research (research group) - https://www.microsoft.com/en-us/research/group/cryptography-research ■ Kim Laine (researcher profile) - https://www.microsoft.com/en-us/research/people/kilai *This on-demand webinar features a previously recorded Q&A session and open captioning. This webinar originally aired on August 21, 2019 Explore more Microsoft Research webinars: https://aka.ms/msrwebinars
Podcast 3 | Homomorphic Encryption with Microsoft Open Source "SEAL" Library
41:51

Podcast 3 | Homomorphic Encryption with Microsoft Open Source "SEAL" Library

In this podcast, Kim Laine (Researcher, Microsoft Research) talks about Homomorphic encryption, its use cases and new version of Microsoft “SEAL” (Simple Encryption Arithmetic Library). Microsoft SEAL is an easy-to-use open source (MIT licensed) homomorphic encryption library developed by the Cryptography Research group at Microsoft. It comes with several examples (with extensive code comments!) in both C++ and C# as Kim pointed out during his walkthrough of this library. Timeline: 1. 0:00 - 1:18 | Introductions 2. 1:19 - 7:28 | Homomorphic Encryption overview 3. 7:29 - 10:45 | Use-cases 4. 10:46 - 15:26 | Current challenges 5. 15:27 - 17:34 | Microsoft SEAL overview 6. 17:35 - 21:41 | Quick walkthrough of example in C++ 7. 21:42 - 26:58 | Running the example 8. 26:59 - 37:02 | Detailed walkthrough example in C# 9. 37:03 - 41:50 | Future direction of Microsoft SEAL Resources: 1. Code: https://github.com/microsoft/SEAL 2. Publications: https://www.microsoft.com/en-us/research/project/microsoft-seal/#!publications 3. Installation Video: https://www.microsoft.com/en-us/research/project/microsoft-seal/#!videos
bottom of page