Will AI replace cybersecurity jobs? What the data says in 2026

No, AI is not replacing cybersecurity jobs; it is one of the reasons there are more of them. The U.S. Bureau of Labor Statistics (BLS) projects information security analyst jobs to grow 21% from 2025 to 2035, seven times the average for all jobs, and it lists the increased use of AI among the reasons security demand is rising.

That doesn’t mean security work is untouched. The same agency rates security analysts very high for AI exposure, and AI is already doing the repetitive parts of the job: sorting alerts, drafting reports and writing procedures. This article looks at the official numbers, what AI can and can’t do in a security role task by task, which jobs around security are actually shrinking, and what the honest risk looks like for people just starting out.

The official numbers: very exposed, fast growing

In August 2026 the BLS published AI exposure ratings for 831 occupations alongside its job projections to 2035. The ratings combine five research measures, including real usage data from Anthropic and Microsoft, and split occupations into four roughly equal groups from low to very high exposure. Every computer job in the table below is in the top group. What differs is the growth:

Three things stand out.

Security is the fast lane in tech. About 192,900 people work as information security analysts in the U.S., and the BLS expects about 233,400 by 2035, with roughly 14,100 openings a year. Of all 831 occupations, only 11 are projected to grow faster. Among the 206 occupations rated very high for AI exposure, only data scientists and computer research scientists grow faster.

AI is part of the reason. The BLS says cyberattacks have become more frequent and that the increased use of AI and the rise of e-commerce have raised the need for security. Every new AI system a company adopts is something else to protect.

The pressure is on routine IT, not on security. Help desk and systems administrator jobs are projected to shrink. Those are roles where AI troubleshooting and automated infrastructure overlap most with the daily work, and they are also common stepping stones into security, which matters for the early-career section below.

The pay reflects the demand. Information security analysts earned a median $129,180 in May 2025, compared with $109,470 for all computer occupations and $50,980 for all jobs.

Why AI creates more security work

Most jobs face AI only as a tool that might do their tasks. Security faces it twice: as a tool for defenders, and as a new attack surface and a new weapon for attackers.

The best numbers on this come from IBM’s 2025 Cost of a Data Breach report, published in July 2025:

• Attackers are using AI. Attackers used AI in 16% of the breaches IBM studied.
• Unmanaged AI is costly. Organizations with high levels of “shadow AI,” tools staff use without approval, saw about $670,000 more in breach costs than those with little or none. And 63% of breached organizations had no AI governance policy or were still writing one.
• Defenders’ AI pays off. Organizations that used security AI and automation extensively saved an average of $1.9 million per breach and cut the time to identify and contain one by 80 days.

Put together, AI raises both the volume of threats and the value of people who can deploy, govern and check AI tools. That is why employers are still hiring. CyberSeek, a workforce tracker run by CompTIA, Lightcast and NIST’s NICE program, counted 514,359 U.S. job listings for cybersecurity and cyber-heavy IT roles in the 12 months to April 2025, up 12% on the year before.

What AI can and can’t do in a security job

The most useful evidence on day-to-day work comes from Microsoft, which analyzed 200,000 anonymized conversations with its Copilot assistant and scored 785 U.S. occupations on how much of their work AI was used for, and how well. Information security analysts scored 0.22, ranking 205th of 785. That is more exposed than about three in four jobs, but well below help desk specialists (37th) and software developers (120th).

The task-level data shows where the line falls. Among the activities that make up a security analyst’s job, AI did best at writing technical procedures and standards and troubleshooting computer problems: it completed roughly 9 in 10 of those requests and covered a large share of each activity. It covered noticeably less of monitoring systems and testing them, and its lowest completion rate, about 83%, came on coordinating with other people to resolve problems. Requests to actually implement security measures were rare in the data.

In practice, a security analyst’s week splits like this:

• Most exposed: documentation and first-pass analysis. Writing policies, procedures and incident reports, summarizing logs, explaining an unfamiliar piece of malware, drafting phishing-awareness material and answering compliance questionnaires. AI drafts these quickly; an analyst still checks them.
• Partly exposed: monitoring and triage. Sorting thousands of alerts, enriching them with context and flagging the ones that look real. AI tools now do much of the sorting, but someone has to decide what is a real incident and what to do about it. ISC2’s 2025 survey found security professionals expect AI’s biggest positive impact in network monitoring and intrusion detection (60%) and endpoint protection and response (56%).
• Least exposed: judgment, response and people. Leading an incident at 2 a.m., deciding whether to take a system offline, weighing a risk against a business deadline, briefing executives, working with legal and regulators, and persuading colleagues to change how they work. These are high-stakes calls that someone has to own.

There is also a reason specific to security: AI tools make confident mistakes, and in security a confident mistake can be a missed breach. That keeps humans in the loop for anything that matters.

How security teams are already using AI

Adoption is real but still cautious. In ISC2’s 2025 AI Pulse Survey of 436 security professionals, published in July 2025, 30% said they had already integrated AI security tools into their operations and 42% were exploring or testing them. ISC2 describes the tools as helping by automating repetitive tasks.

CyberSeek’s data shows employers are only beginning to ask for these skills: about 10% of employers recruiting for cybersecurity roles cited AI as a requirement in the year to April 2025. That gap is an opportunity for anyone willing to learn the tools now.

Which cybersecurity jobs are most and least at risk?

Most exposed:

• Tier-one security operations (SOC) analysts whose day is mostly alert triage and ticket notes. This is where AI tools are being aimed first.
• Help desk and IT support roles, the traditional route into security. Microsoft ranks user support specialists 37th of 785 for AI applicability, and the BLS projects the occupation to shrink 3.5%.
• Compliance-only and documentation-heavy roles that mostly fill in questionnaires, write policies and map controls to frameworks. AI handles much of that drafting well.

Least exposed:

• Incident responders and threat hunters, who investigate live attacks and make decisions under pressure.
• Security architects and cloud security engineers, who design systems and own the trade-offs.
• Security leaders, including CISOs. IT manager jobs are projected to grow nearly 16% by 2035.
• Specialists who secure AI itself, from AI governance to testing AI models for weaknesses, a field that barely existed a few years ago.

Our own scoring tells the same story. Our index rates a typical cybersecurity analyst at 40 out of 100, using the same method our free test applies to a real person’s week. That is well below the most exposed office and tech jobs; a typical software developer week scores 70:

If you’re choosing a direction, these careers from our index combine security or adjacent skills with strong demand:

Will AI take entry-level cybersecurity jobs?

This is the honest weak spot. In the same ISC2 survey, 52% of professionals said AI will significantly or somewhat reduce the need for entry-level staff. The picture is mixed, though: 44% said their organization’s hiring had not yet been affected, and 28% expected AI to create new opportunities for entry-level talent.

The wider evidence points the same way. Anthropic’s March 2026 study of AI and the labor market found tentative evidence that hiring of workers aged 22 to 25 into AI-exposed occupations has slowed slightly, while unemployment for young workers in those jobs stayed flat. We saw the same pattern for junior software engineers.

For someone starting out, the practical lesson is that the old path, a few years on the help desk, then a junior SOC seat, then a specialty, is getting narrower at the bottom. What helps:

• Show hands-on work, such as home labs, capture-the-flag results, write-ups of investigations, or detection rules you built and tested.
• Pair security with a second skill that AI doesn’t commoditize: cloud, networking, scripting or a specific industry such as healthcare or finance.
• Learn the AI tools employers are adopting, and how to secure them. Only about one employer in ten asks for AI skills today, so it still sets you apart.

Is cybersecurity safe from AI? Both sides

The question comes up constantly on Reddit and in career forums, and there are real arguments on both sides.

Why some people think AI will shrink cybersecurity jobs

• AI tools can triage alerts and write reports far faster than a junior analyst.
• Security vendors are building “autonomous SOC” products aimed directly at tier-one work.
• Help desk and sysadmin roles, the usual way in, are projected to shrink.
• A majority of professionals in ISC2’s survey expect fewer entry-level hires.

Why most of the evidence says it won’t

• The official projections show 21% growth for security analysts, among the fastest of any job.
• AI expands what has to be defended and gives attackers new tools, so the work grows with AI rather than shrinking.
• Security decisions carry legal and financial accountability, and organizations want a person responsible for them.
• AI makes confident mistakes, and in security those are expensive. Breaches already cost U.S. organizations a record $10.22 million on average in IBM’s 2025 report.

On balance, cybersecurity is one of the better-placed fields in tech. The risk is concentrated in routine, entry-level and documentation-heavy work, not in the profession.

Will AI replace cybersecurity in the future?

Not on any timeline the official data covers. The BLS projections run to 2035 and show security analysts and IT managers among the fastest-growing jobs in the economy. Beyond that nobody can be sure, but the basic dynamic is hard to break: as long as AI creates new systems, new data and new ways to attack, someone has to secure them.

What will change is the shape of the job. Over the next ten years, expect AI to take most first-pass alert sorting, report writing and routine compliance work, and expect analysts to spend more time investigating, hunting, designing defenses and governing AI tools. Security teams may do more with fewer people per alert, while the number of things to defend keeps rising.

How to AI-proof a cybersecurity career

The rule is the same in every job we cover: let AI do the production and keep ownership of the judgment. Here is what that looks like for each type of task our test measures:

For security specifically, five moves matter most:

1. Move from triage to investigation. The more of your week is spent deciding and responding rather than sorting, the safer it is.
2. Learn to secure AI. Prompt injection, data leakage, model access controls and AI governance are new problems with few experts.
3. Use AI tools daily, and check them. Being the person who knows where the tools fail is worth more than being the one they replace.
4. Build depth in a growing specialty. Cloud security, incident response, identity and security architecture are harder to automate than general monitoring.
5. Practice explaining risk. Briefing executives and persuading colleagues are the least automatable parts of the job, and they lead to leadership roles.

For a broader plan, read our guide to the skills AI can’t replace. To see which jobs really are shrinking, see what jobs will AI replace?, and for the safest careers overall, our ranking of the 25 most AI-proof jobs.