AI is reshaping security jobs, and the first cuts are landing at the bottom
In 2026, AI moved from a talking point in security hiring to a reason given for layoffs. Cloudflare, Arctic Wolf and others tied job cuts to AI investment, and surveys of security leaders now show where the changes are landing: on the analyst roles that have long been the industry's front door.
What vendors are doing
Cloudflare said in May it would cut about 1,100 jobs, roughly a fifth of its workforce, and redesign its processes around AI, according to GovInfoSecurity. Days later, Arctic Wolf, which employs hundreds of people in Canada, cut about 250 roles to fund its AI platform and an "agentic" security operations centre. The cuts at both companies fell largely on commercial and support functions rather than on analysts.
Not every cut is about AI. When Rapid7 announced in August that it would cut 12 per cent of its staff, about 300 jobs, the Boston Globe reported the main pressures were shrinking revenue, customers delaying moves to its newer products and larger rivals gaining ground. Competition from well-funded AI security startups was part of the backdrop.
What surveys show
The 2026 SANS | GIAC workforce report, based on 947 security and hiring leaders, found 74 per cent say AI is affecting team size or structure. But only 16 per cent reported headcount cuts. Far more cited reduced manual analysis time (49 per cent) or workflow automation (48 per cent).
Where roles have changed, the reductions fell most on SOC analysts (32 per cent), threat intelligence analysts (26 per cent) and incident responders (22 per cent), according to SANS. Respondents also reported new roles for AI and machine learning security specialists (34 per cent), AI security engineers (32 per cent) and AI governance analysts (30 per cent).
A separate SANS AI survey in July found 73 per cent of practitioners said AI changed their team's training requirements this year, up from 51 per cent in 2025. It also found that oversight and integration duties are being added to existing roles rather than creating new positions. Only 27 per cent described their AI deployment as mature production, and 63 per cent reported significant AI shortcomings in detection and response.
ISC2's 2025 workforce study found AI was the most-cited skills need, ahead of cloud security, while 72 per cent of respondents agreed that cutting security staff significantly raises breach risk.
Reading the evidence
Three conclusions are reasonable from the data so far. First, AI is changing what security staff do faster than it is eliminating their jobs. Most organizations report efficiency gains, not headcount reduction, and many say the tools still miss things a human analyst would catch.
Second, the pressure is concentrated at the entry level. First-line alert triage is the work AI tools are built to absorb. That puts students and career changers in a harder spot, at the same time as Canadian job-posting data shows entry-level roles stuck at about four per cent of openings.
Third, vendors are making a bet. Companies cutting staff to fund AI products are wagering that automated security operations will win customers. If that bet pays off, demand will rise for people who can build, tune and govern those systems, and fall for people who only work queues.
What it means for practitioners
The SANS researchers put it bluntly: practitioners who use AI are likely to replace those who do not. The practical response is to build skills that sit on top of automation, such as investigating what tools escalate, writing detections, validating AI output and understanding how AI systems themselves can be attacked. New credentials such as CompTIA's SecAI+ and ISACA's AI-focused certifications are early attempts to measure those skills.
For employers, the risk is a hollowed-out pipeline. Senior analysts start as junior ones. Organizations that cut entry-level hiring now may find the experienced staff they need are scarce later.
