A team of researchers has created a ground-breaking algorithm for perfectly secure communications using steganography.
For the first time, ‘perfectly secure’ hidden communications are now possible thanks to research breakthroughs. The technique hides one piece of content inside another in an imperceptible way using recent developments in information theory.
A team of researchers has created a ground-breaking algorithm for perfectly secure communications using steganography, the practise of concealing sensitive information within unimportant content.
The algorithm is a useful tool in digital human communications like social media and private messaging because it can effectively conceal sensitive information so that it cannot be seen that something has been hidden.
The algorithm’s ability to transmit information in a completely secure manner, according to the researchers, may give vulnerable groups like dissidents, investigative journalists, and humanitarian aid workers more power.
Along with additional applications in data compression and storage, this could have significant effects on information security. Researchers have made a significant advancement in secure communications by creating an algorithm that effectively hides sensitive information, making it impossible to tell that anything has been hidden.
This method may soon be widely used in digital human communications, including social media and private messaging, according to the team, which is led by the University of Oxford and works closely with Carnegie Mellon University. Particularly vulnerable groups, like dissidents, investigative journalists, and aid workers, could be strengthened by the ability to send information in an entirely secure manner.
The algorithm is used in steganography, which is the practise of concealing sensitive information within seemingly unimportant content. Steganography is different from cryptography in that it hides sensitive information in a way that makes it difficult to detect its concealment. An illustration would be to conceal a Shakespearean poem within an AI-created cat image.
Existing steganography techniques have been studied for more than 25 years, but they typically have weak security, making it possible for users to be discovered. This is due to the fact that earlier steganography algorithms would slightly alter how the harmless content was distributed.
The research team overcame this by utilising recent developments in information theory, particularly minimum entropy coupling, which enables the joining of two data distributions in a way that maximises their mutual information while preserving the individual distributions.
Therefore, the distribution of harmless content and the distribution of content that encodes sensitive information do not differ statistically under the new algorithm. The algorithm was tested using a variety of models that create automatically generated content, including the text-to-speech converter WAVE-RNN and the open-source language model GPT-2.
The new algorithm demonstrated up to 40% higher encoding efficiency than earlier steganography techniques across a range of applications in addition to being completely secure, allowing for the concealment of more information within a given amount of data. Due to the advantages for data compression and storage, steganography may become a desirable technique even if perfect security is not required.
The research team has filed a patent for an algorithm, which they intend to issue under a free license to third parties for non-commercial responsible use. They have published their work as a preprint paper on arXiv and open-sourced an inefficient implementation of their method on Github.
They will present the algorithm at the 2023 International Conference on Learning Representations in May. AI-generated content is increasingly used in ordinary human communications, and steganography may become more widespread.
Co-lead author Dr. Christian Schroeder de Witt (Department of Engineering Science, University of Oxford) developed a method that can be applied to any software that automatically generates content.
This could be useful for journalists and aid workers in countries where encryption is illegal, but users should be aware of side-channel attacks such as detecting a steganography app on the user’s phone.
Co-lead author Samuel Sokota and contributing author Professor Jakob Foerster have developed a new family of steganography algorithms that have perfect security guarantees.
The study involved Prof. Zico Kolter at Carnegie Mellon University, USA, and Dr. Martin Strohmeier from armasuisse Science+Technology, Switzerland. The work was partially funded by an EPSRC IAA Doctoral Impact Fund hosted by Professor Philip Torr at the University of Oxford.