Iron Man wouldn’t be as smooth without Jarvis, his AI assistant. He manages everything in Tony Stark’s life and warns him when something is about to go wrong.

IBM Security has built an artificial intelligence system to warn agencies of digital weaknesses. It doesn’t just answer specific questions, but constantly scans for information, understands what it means, and makes recommendations to tackle new challenges – which could otherwise be unknown.

A tweet or a blog post could hold the clue to stopping a cyber attack. The AI system can “read” books, reports and blogs, “see” images, and “hear” people speak to warn in advance of an attack.

Such information – expressed in natural human language – makes up 80% of all data in the world, and was previously inaccessible for computer analysis. A traditional security system wouldn’t be able to make sense of such information, and it would take too long for a human to scan all of social media.

To find out more about how AI can be used in digital security, download the full report by IBM: