Does artificial intelligence risk putting hackers out of work?

AI does not spell the end of hackers, it redefines their profession.

The meteoric rise of generative artificial intelligence is disrupting all sectors, including cybersecurity. A burning question is agitating minds: are hackers, whether ethical (white hats) or malicious (black hats), at risk of finding themselves unemployed in the face of ever more powerful AI? The answer is nuanced. Far from signaling their disappearance, AI is profoundly transforming their methods, their tools and the skills required. This article analyzes the real impact of AI on the hacker job market, drawing on recent data and concrete examples.

AI, a new tool for hackers: automation and amplification

Artificial intelligence is not just a threat to hackers; it is also a formidable tool that they massively adopt.

Massive and rapid adoption

The numbers speak for themselves. According to a 2026 Bugcrowd study, 82% of hackers and security researchers use AI tools in their workflow daily, up from 71% a year earlier. This rapid adoption shows that AI has become a central part of their arsenal, whether for reconnaissance, vulnerability analysis or code generation.

Increased offensive capabilities

AI makes it possible to automate tasks that were previously time-consuming and complex. Generative models can produce code, analyze configurations, test operating hypotheses, and construct attack scenarios. Combined with agents capable of executing actions on real systems, they make it possible to orchestrate offensive operations at a speed impossible for a human team.

A striking example is that of AI-boosted “script kiddies”. According to Palo Alto Networks' Unit 42 firm, these low-skilled hackers now have "the same tools and the same level of sophistication as a state-backed group." In three weeks of internal testing, Unit 42 carried out the equivalent of one to two years of intrusion testing using dedicated AI models.

Speed ​​as a new weapon

One of the major changes brought about by AI is the compression of time. Where a sophisticated intrusion previously required several days or weeks, certain steps can now be automated at machine speed. Cybersecurity is entering a phase where the attack can progress at machine speed, forcing defenders to rethink their strategies.

AI does not yet replace hackers: limits persist

Despite these spectacular advances, AI is far from being able to do without human expertise. The data even shows a growing distrust of the total autonomy of AI when it comes to offensive security.

Confidence at half mast

Telling fact: Although 82% of hackers use AI daily, trust in fully autonomous AI penetration testing has fallen from 29% to just 9% in one year. Organizations and professionals are realizing that AI alone is not yet reliable to carry out a complex end-to-end attack.

LLMs fail cybersecurity benchmarks

State-of-the-art language models (LLMs) are not yet standalone hackers. Simbian tested 11 state-of-the-art LLMs on 105 attack procedures and 884 executions. None succeeded. The best model, Claude Opus 4.6, only achieved an average performance of 46% on MITER tactics and found only 4.5% of malicious indicators. Simbian's conclusion is clear: “an LLM alone is not a defender”. The same goes for the attack.

The need for human context

AI models are excellent at discovering vulnerabilities repetitively, but they remain weaker than human hackers at understanding context, particularly in identifying the most valuable information in an organization. A human hacker knows how to judge whether a strange connection is the work of an intern, an administrative script, or the start of a lateral movement. This ability to judge under pressure remains difficult to automate.

The real impact on employment: shortage and change, not disappearance

Contrary to popular belief, AI is not destroying cybersecurity jobs. It transforms them and even creates strong demand.

A global talent shortage

The cybersecurity sector faces a chronic labor shortage. ISC2 estimates the global cybersecurity professional gap at 4.76 million people, up 19.1% year-on-year. A profession that is short of millions of people is not about to become overcrowded.

A growth in job offers

Labor market data confirms this trend. According to Glassdoor, cybersecurity job postings increased 11% in the first quarter of 2026 compared to the same period in 2025. Companies aren't cutting staff; on the contrary, they seek to recruit more to face new threats.

New roles emerge

AI creates new specific professions. We are seeing positions like “AI Red Team Specialist”, “AI/ML Security Engineer” or “AI Threat Intelligence Analyst”, with salaries between $110,000 and $240,000 per year. The number of cybersecurity job postings requiring AI skills has doubled in one year across G7 countries, rising to 28.5% of positions.

The evolution of skills

AI does not eliminate jobs, it changes the organization of work. Repetitive tasks such as first-level triage or basic reports are automated, but this frees professionals for higher value-added missions: judgment, strategy, crisis management. New recruits can therefore move more quickly from routine tasks to strategic interventions.

Hackers are also worried about AI

Unexpected phenomenon: cybercriminals themselves express fears about the rise of AI. Sophos researchers analyzed discussions on dark web forums and underground marketplaces. Many malicious hackers worry about losing their jobs to AI.

Some fear that AI tool vendors will take the work of those who manually write their malicious scripts. Others point out that the quality of “products” (malware) could deteriorate if developers outsource their coding to AI. However, these forums remain divided, and some members accuse companies of “overreacting” to the potential of models like Claude Mythos.

Conclusion: a transformation, not a disappearance

Artificial intelligence will not put hackers out of work. It radically redefines their profession. Hackers who can leverage AI to automate repetitive tasks and focus on strategic aspects – understanding context, decision-making under pressure, offensive creativity – will remain essential. Those who cling to purely manual methods risk being marginalized.

The real question is therefore not “will AI replace hackers?” ", but "how will hackers adapt to AI? ". For cybersecurity professionals, the answer is clear: you must learn to work with AI, develop dual technical and strategic skills, and cultivate human qualities that machines cannot yet imitate. The future belongs to those who know how to team up with artificial intelligence, rather than fearing it.