Why AI Is The Future Of Cybersecurity
- 61% of enterprises say they cannot detect breach attempts today without the use of AI technologies.
- 48% say their budgets for AI in cybersecurity will increase by an average of 29% in Fiscal Year (FY) 2020.
- Breach attempts are proliferating with Cisco reporting that in 2018, they blocked seven trillion threats on behalf of their customers.
These and many other insights are from Capgemini’s Reinventing Cybersecurity with Artificial Intelligence Report published this week. You can download the report here (28 pp., PDF, free, no opt-in). Capgemini Research Institute surveyed 850 senior executives from seven industries, including consumer products, retail, banking, insurance, automotive, utilities, and telecom. 20% of the executive respondents are CIOs, and 10% are CISOs. Enterprises headquartered in France, Germany, the UK, the US, Australia, the Netherlands, India, Italy, Spain, and Sweden are included in the report. Please see page 21 of the report for a description of the methodology.
Capgemini found that as digital businesses grow, their risk of cyberattacks exponentially increases. 21% said their organization experienced a cybersecurity breach leading to unauthorized access in 2018. Enterprises are paying a heavy price for cybersecurity breaches: 20% report losses of more than $50 million. Centrify’s most recent survey, Privileged Access Management in the Modern Threatscape, found that 74% of all breaches involved access to a privileged account. Privileged access credentials are hackers’ most popular technique for initiating a breach to exfiltrate valuable data from enterprise systems and sell it on the Dark Web.
Key insights include the following:
- 69% of enterprises believe AI will be necessary to respond to cyberattacks. The majority of telecom companies (80%) say they are counting on AI to help identify threats and thwart attacks. Capgemini found the telecom industry has the highest reported incidence of losses exceeding $50M, making AI a priority for thwarting costly breaches in that industry. It’s understandable by Consumer Products (78%), and Banking (75%) are 2nd and 3rd given each of these industry’s growing reliance on digitally-based business models. U.S.-based enterprises are placing the highest priority on AI-based cybersecurity applications and platforms, 15% higher than the global average when measured on a country basis.
- 73% of enterprises are testing use cases for AI for cybersecurity across their organizations today with network security leading all categories. Endpoint security the 3rd-highest priority for investing in AI-based cybersecurity solutions given the proliferation of endpoint devices, which are expected to increase to over 25B by 2021. Internet of Things (IoT) and Industrial Internet of Things (IIoT) sensors and systems they enable are exponentially increasing the number of endpoints and threat surfaces an enterprise needs to protect. The old “trust but verify” approach to enterprise security can’t keep up with the pace and scale of threatscape growth today. Identities are the new security perimeter, and they require a Zero Trust Security framework to be secure. Be sure to follow Chase Cunningham of Forrester, Principal Analyst, and the leading authority on Zero Trust Security to keep current on this rapidly changing area. You can find his blog here.
- 51% of executives are making extensive AI for cyber threat detection, outpacing prediction, and response by a wide margin. Enterprise executives are concentrating their budgets and time on detecting cyber threats using AI above predicting and responding. As enterprises mature in their use and adoption of AI as part of their cybersecurity efforts, prediction and response will correspondingly increase. “AI tools are also getting better at drawing on data sets of wildly different types, allowing the “bigger picture” to be put together from, say, static configuration data, historic local logs, global threat landscapes, and contemporaneous event streams,” said Nicko van Someren, Chief Technology Officer at Absolute Software.
- 64% say that AI lowers the cost to detect and respond to breaches and reduces the overall time taken to detect threats and breaches up to 12%. The reduction in cost for a majority of enterprises ranges from 1% – 15% (with an average of 12%). With AI, the overall time taken to detect threats and breaches is reduced by up to 12%. Dwell time – the amount of time threat actors remain undetected – drops by 11% with the use of AI. This time reduction is achieved by continuously scanning for known or unknown anomalies that show threat patterns. PetSmart, a US-based specialty retailer, was able to save up to $12M by using AI in fraud detection from Kount. By partnering with Kount, PetSmart was able to implement an AI/Machine Learning technology that aggregates millions of transactions and their outcomes. The technology determines the legitimacy of each transaction by comparing it against all other transactions received. As fraudulent orders were identified, they were canceled, saving the company money and avoiding damage to the brand. The top 9 ways Artificial Intelligence prevents fraud provides insights into how Kount’s approach to unsupervised and supervised machine learning stops fraud.
- Fraud detection, malware detection, intrusion detection, scoring risk in a network, and user/machine behavioral analysis are the five highest AI use cases for improving cybersecurity. Capgemini analyzed 20 use cases across information technology (IT), operational technology (OT) and the Internet of Things (IoT) and ranked them according to their implementation complexity and resultant benefits (in terms of time reduction). Based on their analysis, we recommend a shortlist of five high-potential use cases that have low complexity and high benefits. 54% of enterprises have already implemented five high impact cases. The following graphic compares the recommended use cases by the level of benefit and relative complexity.
- 56% of senior execs say their cybersecurity analysts are overwhelmed and close to a quarter (23%) are not able to successfully investigate all identified incidents. Capgemini found that hacking organizations are successfully using algorithms to send ‘spear phishing’ tweets (personalized tweets sent to targeted users to trick them into sharing sensitive information). AI can send the tweets six times faster than a human and with twice the success. “It’s no surprise that Capgemini’s data shows that security analysts are overwhelmed. The cybersecurity skills shortage has been growing for some time, and so have the number and complexity of attacks; using machine learning to augment the few available skilled people can help ease this. What’s exciting about the state of the industry right now is that recent advances in Machine Learning methods are poised to make their way into deployable products,” said Nicko van Someren, Chief Technology Officer at Absolute Software.
AI and machine learning are redefining every aspect of cybersecurity today. From improving organizations’ ability to anticipate and thwart breaches, protecting the proliferating number of threat surfaces with Zero Trust Security frameworks to making passwords obsolete, AI and machine learning are essential to securing the perimeters of any business. One of the most vulnerable and fastest-growing threat surfaces are mobile phones. The two recent research reports from MobileIron, Say Goodbye to Passwords (4 pp., PDF, opt-in) in collaboration with IDG, and Passwordless Authentication: Bridging the Gap Between High-Security and Low-Friction Identity Management (34 pp., PDF, opt-in) by Enterprise Management Associates (EMA) provide fascinating insights into the passwordless future. They reflect and quantify how ready enterprises are to abandon passwords for more proven authentication techniques including biometrics and mobile-centric Zero Trust Security platform.