AI has high expectations in data centers, but its potential is still unproven
AI is becoming increasingly popular in data centers, but its full potential is still unproven. There are both pros and cons to using AI in data centers, and organizations need to carefully consider their needs before investing in this technology.
On the one hand, AI can be used to improve security, energy efficiency, and physical security. For example, AI can be used to identify and respond to threats, optimize energy usage, and prevent unauthorized access.
On the other hand, AI can also be expensive to implement and maintain. Additionally, there are concerns about the security and privacy of AI-powered systems.
Overall, AI is a promising technology with the potential to revolutionize data centers. However, organizations need to be aware of the challenges and risks before making the leap to AI.
Specifically, define and clearly articulate the desired results.
Hoff suggests defining expected outcomes, including data storage and data lifecycle management. Without clear expectations and an understanding of the input, judging the output's efficacy is difficult. Incomplete definitions and outputs may lead to cybersecurity threats slipping through.
Security professionals sometimes misunderstand AI's capabilities and cost. Some expect AI to drastically cut costs by replacing staff, but this is unlikely, says Mauricio Sanchez, senior director of market research at Dell'Oro Group.
Hoff believes that human analysts will continue reviewing and validating crucial decisions made by AI. While AI can handle data collection, human judgment is necessary to ensure the AI's accuracy. Trusting machines completely will take time, according to Hoff.
Make sure to do your research before investing in AI.
Users are analyzing whether to invest in first-generation AI technology or wait for improved capabilities, according to Sanchez. CISOs should assess if AI can enhance SOC capabilities, reduce risk, and offer a positive ROI. Buying solely on faith is not recommended.
AI features are often offered through subscriptions, meaning clients must know how to utilize the service effectively. Vendors aim for adoption and usage, not just selling subscriptions, to ensure customer satisfaction, as Boujelbene explains.
Sanchez explained the technology curve and the challenges of first-generation solutions. New technologies may require multiple attempts before successful adoption.
AI can greatly impact data center management, automating tasks and optimizing performance, according to Chaudhuri from Dasera. However, integrating AI also poses risks, such as flawed decision-making due to biased training data and over-reliance on AI systems. A balanced approach with human oversight is crucial.