Network segmentation, MCP, and how to make AI trustworthy
Segmenting a network might keep AI away from critical data, but protocols like MCP could threaten a tech infrastructure entirely.

Caroline Nihill is a reporter for IT Brew. Her work primarily focuses on cybersecurity, IT operations, and tech interactions between the private and public sectors. Based in the D.C. area, Caroline has also worked for Scoop News Group outlets including FedScoop, CyberScoop, StateScoop, and others.
Segmenting a network might keep AI away from critical data, but protocols like MCP could threaten a tech infrastructure entirely.
The organization reported an 89% increase in AI-enabled adversary attacks year-over-year.
While cyberattackers may try to take advantage of the war in the Middle East, one expert says there's no major escalation at the moment.
One leader says it’s a business risk if a technologist doesn’t have a say in revenue discussions.
One expert reported seeing more attackers going after third parties for data.
One expert says that professionals can expect to see more extortion attempts as attackers monetize data from past breaches.
“It’s like when you’re trying to walk through a muddy area or something, you’re going to get dirty at some point,” Kaseya CISO says.
Experts want those in IT to know about preparing for a potential ransomware attack, and organizations’ risk levels.
“It’s Whac-a-Mole, or a game of cat and mouse, between defenders and attackers,” one cybersecurity leader says.
“What is the risk of adopting it, cost of adopting it, and do we actually need it?” one expert asks.