A new AI-based system from Nvidia sniffs out unusual behavior and ties it to users, in an effort to prevent insider attacks and protect digital credentials. Credit: Rick Jo / Getty Images Nvidia today announced that a digital lab playground for its latest security offering is now available, letting users try out an AI-powered system designed to monitor individual user accounts for potentially hazardous behavior.The idea, according to the company, is to leverage the large amounts of data that many organizations compile anyway about login and data access events on their systems, and use that to train an AI that watches for user accounts to diverge from their usual patterns. The system moves security teams from a scenario in which they have to comb through potentially millions of events a week to identify a problem to a small handful of “high risk” events identified by the system.A digital fingerprinting “LEGO kit”In an official blog post, Nvidia said that the solution is essentially “a kind of LEGO kit” for digital fingerprinting, allowing users to customize their own solution that best meets their needs. Justin Boitano, Nvidia vice president of enterprise and edge computing, said in a briefing that the system has the potential to both reduce workloads and enable faster identification of potential malefactors. “This digital fingerprinting API workflow enables security organizations to detect threats instantly,” he said. “And this means identity attacks, which are the most common form of enterprise security attacks where bad actors are going after employee credentials via social engineering, or phishing attempts, can quickly get mitigated and detected instantly.”Nvidia said that systems like this are just the beginning for AI-backed security efforts. In future, more detailed user profiles, including data as granular as typing speed and accuracy, could be used to help identify suspicious behavior, the company said in its blog post. “Data on network events is gold for building AI models that harden networks, but no one wants to share details of real users and break-ins,” Nvidia wrote. “Synthetic data, generated by a variant of digital fingerprinting, could fill the gap, letting users create what they need to fit their use case.”The digital fingerprinting system is currently available in Nvidia’s AI Enterprise platform, and the new lab module – which allows users to walk through a real deployment on sample architecture hosted at Equinix and interact with Nvidia experts – is available via the company’s LaunchPad website as of today. Related content opinion The Assumed Breach conundrum Assumed Breach is the third but often overlooked principle of zero trust. When we talk about adopting a “not if, but when” attitude to security, are we merely paying lip service or do we really believe and internalise it? By Steven Sim Apr 23, 2024 4 mins Zero Trust Security news Authentication failure blamed for Change Healthcare ransomware attack Absence of multi-factor authentication reportedly left a remote access application exposed. By John Leyden Apr 23, 2024 5 mins Ransomware Cyberattacks news Russian state-sponsored hacker used GooseEgg malware to steal Windows credentials A now-patched Windows Print Spooler flaw was used by Forest Blizzard to drop the privilege-elevating malware for credential stealing and persistence. By Shweta Sharma Apr 23, 2024 3 mins Malware Windows Security feature Top 10 physical security considerations for CISOs Securing premises and devices from physical attacks can be just as challenging as defending against cyber threats. Collaboration and communication with all teams involved is the key to success. By Ericka Chickowski Apr 23, 2024 14 mins Critical Infrastructure Security Infrastructure Security PODCASTS VIDEOS RESOURCES EVENTS SUBSCRIBE TO OUR NEWSLETTER From our editors straight to your inbox Get started by entering your email address below. Please enter a valid email address Subscribe