Corporations that are attempting to ward off hackers are beginning to use Artificial Intelligence (AI) to sort through a vast array of malware files to identify common characteristics that will help them prevent new attacks. Machine learning is also assisting in analyzing people’s voices, fingerprints and typing styles to make sure that only authorized users get into systems that have accounts and customer data stored – this is particularly important to banks, brokerage houses, and other entities where customers want to access and manipulate their accounts, but want no one else to be able to do that. As an additional assistance in the fight for cybersecurity, John Bolton, National Security Adviser, has announced that an earlier order which effectively tied the federal government’s hands in combating cybersecurity invasion had been rescinded. The new process allows the Defense Department more flexibility in launching offensive cyberstrikes without first having to vet those decisions through an elaborate interagency process. Our tech support team has indicated that the changes have already begun to make a big difference in the number of security incursions. Other efforts that are currently at work in large corporations seek to combat zero-day malware from running on their systems – that is, threats that are unknown to the security community at the time it appears. AI is helping to solve this problem and identify new attacks as soon as they appear. The systems analyze existing malware and determine what characteristics the files have in common, then check to see if potential new threats have any of those traits. Much of the malware has, in the past, gained entry into a company’s system by an employee clicking on a malicious file. Thus, using the new systems, when a user in the company clicks on a suspicious file, the company’s tool scans hundreds of different attributes, then runs them through a machine-learning (AI) algorithm that compares them to the company’s malware database and determines how likely the file is to be malicious. Machine learning works well for malware because AI can readily attack the vast amount of data out there and spot problems. One current difficulty being encountered is the “false positives” where benign files are identified as malicious. But misidentification in that direction seems a small price to pay for sifting through the many reams of data in a concerted effort to stamp out malicious malware. Here’s hoping that all companies that retain databases of customer information are participating in this kind of concerted effort, so that there won’t be a repeat of the situations of the past few years where vast amounts of customers’ personal information was lost to hackers.