Artificial intelligence systems can create fake documents to fool enemies.
British intelligence agents created false documents on a corpse to fool Nazi Germany and convince them that they were preparing to attack Greece during World War II. Operation Mincemeat was a success and covered the actual Allied invasion of Sicily.
The “canary trap” technique spreads false documents in multiple versions to hide a secret in espionage. Canary traps detect information leaks or create distractions to hide valuable information.
WE-FORGE is a new data protection system developed in the Department of Computer Science. It uses artificial intelligence to expand on the canary trap concept. The system automatically creates fake documents to protect intellectual property, such as military technology and drug design.
V.S. says, “The system produces documents which are sufficiently similar to their original to be plausible but sufficiently different from each other to be inaccurate.” Subrahmanian, the Distinguished Professor of Cybersecurity, Technology, and Society and Director of the Institute for Security, Technology, and Society, is Subrahmanian.
Cybersecurity professionals use canary traps or “honey files” to deceive potential attackers. Foreign language translators can also be used. WE-FORGE uses natural language processing to generate multiple fake files that can be believed or false. The system adds a random element to prevent adversaries from identifying the document.
WE-FORGE is a new data protection system developed in the Department of Computer Science. It uses artificial intelligence to expand on the canary trap concept. The system automatically creates fake documents to protect intellectual property, such as military technology and drug design.
V.S. says, “The system produces documents which are sufficiently similar to their original to be plausible but sufficiently different from each other to be inaccurate.” Subrahmanian, the Distinguished Professor of Cybersecurity, Technology, and Society and Director of the Institute for Security, Technology, and Society, is Subrahmanian.
Cybersecurity professionals use canary traps or “honey files” to deceive potential attackers. Foreign language translators can also be used. WE-FORGE uses natural language processing to generate multiple fake files that can be believed or false. The system adds a random element to prevent adversaries from identifying the document.
WE-FORGE can create multiple fake versions of any technical design documents. Hackers are often faced with difficulty figuring out which of many similar documents is genuine when they hack into a system.
“Using this technique, we force adversaries to waste their time and effort identifying correct documents. Subrahmanian says that even if they succeed, they might need more confidence to do it correctly.
It is not easy to create false technical documents. The research team estimates that a single patent could include over 1,000 concepts and up to 20 replacements. In one technical paper, WE-FORGE can consider millions of possible replacements for the concepts.
Subrahmanian says that “malicious actors” are now stealing intellectual property and getting away with it free of charge. This system increases thieves’ costs when they steal secrets from government agencies or industries.
The WE-FORGE algorithm uses similarities to find words in documents and analyzes how they relate. The system sorts the concepts into “bins” and then computes the most likely candidate for each group.
Dongkai Chen, Guarini 21 says WE-FORGE can also get input from the original author. “The combination of human and machine ingenuity can increase costs on intellectual-property thieves even more.”
The research team falsified several computer science and chemical patents and then asked a panel to determine which documents were authentic.
The research published in ACM Transactions on Management Information Systems found that WE-FORGE could ” consistently generate highly believable fake documents to each task.”
WE-FORGE is a tool that specializes in the falsification of technical information. It does not hide simple details like passwords, unlike other devices.
WE-FORGE is an improvement on FORGE, the previous version of the system. It eliminates the need to create guides that describe concepts for specific technologies. WE-FORGE ensures greater variety among fakes and uses a better method to select images to replace or their replacement.