Artificial Intelligence (AI) is a technology that we're slowly becoming more exposed to as new products are developed to make everyday tasks easier. AI actually had a bit of an interesting start - it grew from the technology used to break the German Enigma codes in World War II, however we are more familiar with its more benign uses in beating chess champions or even in our virtual phone assistants. Enterprises have been quietly developing AI-based cybersecurity strategies that are proving to have promise in protecting systems and even are being used in counterterrorism efforts. These developments are explored in an article on Computer Weekly.
Using declassified US defense and intelligence strategies, entrepreneur Drew Perez founded Adatos, a Singapore-based AI startup that has developed software to produce data-driven insights quickly. Perez says that machine learning is a popular tool among intelligence services and the US military to process large amounts of data, and advances have made the usefulness of the interpreted data incredible. "The current demands in counter-terrorism require precise, accurate insights delivered in the span of minutes,” he says. “A key contest in war will be between adversary cognitive systems – artificial and human – to process information, understand the battlespace and decide and execute faster than the enemy."
However, Perez cautions that we need not fear that AI become a sentient threat like it is in the movies; "AI, if defined by the expectations of cognitive functions that mimic humans, is still largely in a development stage, but it doesn’t mean it can’t solve real-world problems much more efficiently.”
AI is used in many applications that we may not even notice - in customer services, robo-advisors (think IKEA's "Anna"), retail projections, robotics, and even in interpreting metical data - but the technology is always improving and we are seeing more and more of it at our fingertips. Surveillance systems can be programmed to search for the faces of wanted persons within a feed, and Tesla has even introduced the beginnings of mass produced autonomous cars with its Autopilot feature.
In cybersecurity, machine learning allows systems to recognize known threats and respond to their attacks appropriately. Attackers and their methods can also be filed into a database for future reference, but where this technology falls short is in handling new threats - the system simply does not have the data to draw on. To address this "high-fidelity machine learning" is available, which "runs tests before and during malware execution, allowing security applications to halt malicious operations halfway and kick malware out of the system." High-fidelity machine learning can also reduce false positive, behavioral analysis, and preventing OS exploitation. With automation on board, the system load required to handle security issues can be reduced.
AI has its limitations though - it lacks the ability of deep analysis and data is difficult to process when it exists in silos. Algorithms also vary in accuracy. Ultimately - AI should not be used as a tool to replace people just yet. It lacks the ability to understand the intricacies of human expression and cultural norms and is likely to offend more than help, but these issues improve every day.
Read more on this article at Computer Weekly.