Channel faces challenges around AI security

The rise of artificial intelligence (AI) has both sparked security concerns as well as provided those protecting data with more tools to use to counter cyber criminals.

The double-sided nature of AI, both as a protector and source of risk, has made it an essential technology for the channel to master to support their own operations as well as those of customers.

Vendors, disties and managed service providers (MSPs) have all moved to generate, stock and deliver products and services that improve AI security, but the industry is dealing with several challenges, including speed, volume and skills.

“Artificial intelligence is reshaping cyber security faster than almost any other enterprise technology. Machine learning-driven detection and response tools are helping organisations spot anomalies and threats in near real time, but the same advances are equally available to attackers,” said Richard Eglon, CMO of Nebula Global Services.

Richard Ford, CTO at cyber security specialist Integrity360, was also concerned about the speed with which AI was changing customer environments and exposing them to fresh levels of risk

“Defenders are trying to keep pace by automating detection and response,” he said. “The risk is that organisations adopt AI without the same level of control. If you don’t fully understand what your AI systems can access, or how they behave in real environments, you’re introducing risk at speed.”

The channel also must wrestle with the sheer number of options that are emerging in the booming AI security product segment to work out which produces they should recommend to users.

Sunil Agrawal, CISO at Glean, pointed out that the problem was not a shortage of security products but “a surplus. Too many tools, too few people to run them and none of them talking to each other.”

He believed agentic AI was “the connective tissue” that orchestrated defences autonomously “without needing a small army of trained analysts to babysit every dashboard”, describing it as “a structural shift in favour of defenders”.

Eglon was also concerned abour the strain AI is putting on the levels of expertise available across the channel, with skills always under pressure: “Skills are the other major constraint. Delivering an AI-enabled security platform is very different from deploying a single firewall or endpoint agent. Platforms require tuning, integration with customer environments, ongoing optimisation and a deep understanding of how AI models behave.

“Channel research consistently highlights a cyber security skills gap, particularly around data science, automation and threat hunting. Successful partners are responding by investing in vendor certifications, building security operations centres [SOCs], and using their own automation and managed detection services to spread scarce expertise across multiple customers,” he said.

One of the responses from customer keen to use AI security tools effectively has been to turn to the platform approach, an option being encouraged by vendors.

Rob Anderson, director of security practice for global channel at Barracuda, stated that customers are turning to platforms because they “want to reduce tool sprawl, drive automation, increase visibility and simplify day-to-day operations for under-resourced or over-stretched security teams”. 

Eglon added that the platform approach was not something all vendors could do and the channel needed to be careful in choosing their partners.

“The platform shift also raises uncomfortable questions for vendors that cannot offer a broad suite. I’m sure some will double down on being best-of-breed in a critical niche, relying on strong integration to remain relevant. Others may seek partnerships or identify specific acquisition targets. From a channel perspective, this fragmentation reinforces the need to curate ecosystems carefully and to avoid over-committing to suppliers with a narrow roadmap,” he said.