Will AI replace ethical hackers?
Updated
Penetration testers score 60 of 100 for AI exposure. The largest share of their task time, 73%, is work AI speeds up while they stay in charge. AI can take over 18% outright, and 9% stays with people.
- 18% AI can do it
- 73% AI helps
- 9% Needs a person
High exposure
Higher than 91% of the 923 occupations we scored.
Task by task
We graded 22 tasks, listed here from most to least exposed. AI can do 4 of them today, speeds up 16 and leaves 2 to people.
| Task | AI exposure |
|---|---|
| Write audit reports to communicate technical and procedural findings and recommend solutions. AI drafts audit reports from findings; auditor reviews and signs. | 65% AI can do it |
| Document penetration test findings. AI drafts findings from notes and evidence; tester checks accuracy. | 62% AI can do it |
| Develop presentations on threat intelligence. AI drafts slides and summaries from threat intel content quickly. | 60% AI can do it |
| Gather cyber intelligence to identify vulnerabilities. AI search and summarization speed OSINT; analyst validates findings. | 60% AI helps |
| Prepare and submit reports describing the results of security fixes. AI drafts fix status reports from tickets; reviewer confirms outcomes. | 60% AI can do it |
| Evaluate vulnerability assessments of local computing environments, networks, infrastructures, or enclave boundaries. AI can triage scan results; evaluator judges risk in context. | 55% AI helps |
| Conduct network and security system audits, using established criteria. AI can parse scan output against checklists; auditor verifies results. | 50% AI helps |
| Configure information systems to incorporate principles of least functionality and least access. AI drafts hardening configs; admin must apply and verify them. | 50% AI helps |
| Develop security penetration testing processes, such as wireless, data networks, and telecommunication security tests. AI drafts test procedures and checklists; experts validate method choices. | 50% AI helps |
| Identify new threat tactics, techniques, or procedures used by cyber threat actors. AI summarizes threat reports fast; analyst confirms new tactics. | 50% AI helps |
Show the other 12 tasks
| Task | AI exposure |
|---|---|
| Keep up with new penetration testing tools and methods. AI summarizes new tools and papers; practitioner tests what matters. | 50% AI helps |
| Update corporate policies to improve cyber security. AI drafts policy language; leadership must approve and take responsibility for policy. | 50% AI helps |
| Develop and execute tests that simulate the techniques of known cyber threat actors. AI drafts attack plans and scripts; tester runs live emulation. | 45% AI helps |
| Identify security system weaknesses, using penetration tests. AI assists planning and triage; tester drives the live testing. | 45% AI helps |
| Maintain up-to-date knowledge of hacking trends. AI can track trend reports and summarize; person filters relevance. | 45% AI helps |
| Design security solutions to address known device vulnerabilities. Design needs architecture judgement; AI generates options and checks gaps. | 40% AI helps |
| Develop infiltration tests that exploit device vulnerabilities. AI drafts exploit code; tester adapts it to live targets. | 40% AI helps |
| Investigate security incidents, using computer forensics, network forensics, root cause analysis, or malware analysis. AI aids log and malware triage; analysts take responsibility for root cause. | 40% AI helps |
| Collect stakeholder data to evaluate risk and to develop mitigation strategies. Interviews stay human led; AI helps with prep and summaries. | 35% AI helps |
| Test the security of systems by attempting to gain access to networks, Web-based applications, or computers. Live exploitation needs a skilled tester; AI assists with recon. | 35% AI helps |
| Discuss security solutions with information technology teams or management. Live discussions with IT and management rely on human trust. | 18% Needs a person |
| Assess the physical security of servers, systems, or network devices to identify vulnerability to temperature, vandalism, or natural disasters. Requires walking the site to inspect hardware and facilities. | 10% Needs a person |
How penetration testers compare
All occupations (923)
Computer and mathematical (36)
Pay and outlook in the US
- Median pay
- $116,580 a year, May 2025
- People employed
- 48,374 in the US, May 2025
- Projected change in jobs
- +5.1% from 2024 to 2034
Skills to build
- Adapt AI-drafted exploit code to live targets in Python
- Run live exploitation and privilege escalation on Linux hosts
- Assess AWS accounts for misconfigured IAM roles and buckets
- Verify AI-drafted findings against captured evidence before sign-off
- Hold the OSCP certification from OffSec
- Conduct physical site walks of servers and network devices
Safer roles nearby
5 related occupations with lower scores.
Questions about penetration testers and AI
Will ethical hackers be replaced by AI?
Partly. About 5 in 10 of a penetration tester's working hours go to tasks AI can already do, mostly writing audit reports and documenting test findings. Hands-on exploitation, physical site checks and stakeholder talks stay with the tester, who reviews and signs what AI drafts.
Which tasks are AI taking from penetration testers?
Report writing is taken over first, with audit reports at exposure 65 and penetration test findings at exposure 62. AI drafts audit reports from findings and documents test findings from notes and evidence, while the tester checks accuracy and signs off. Threat intelligence slides are automated too, since AI drafts them from intel content quickly.
Which tasks stay with penetration testers?
Physical site checks stay human. Testers walk the site to inspect servers, network devices and facilities for heat, vandalism or disaster risk. Discussions with IT teams and management also stay human, because they rely on trust built in live conversation. Stakeholder interviews are human led, with AI only helping prepare questions and summaries.
How can penetration testers stay valuable alongside AI?
Keep the live work that needs a person in the room, then check everything AI drafts. Run live exploitation on Linux and UNIX hosts, adapt exploit code in Python and Bash, and walk sites in person. Review every AI-drafted finding against captured evidence before it reaches a client. Testers who show the evidence behind a finding keep the sign-off.
Related jobs
Next steps for penetration testers
Check which roles in your team AI can take on
We map your team's tasks, show what AI can do today and build the automations that pay back first.
Related topics
- Jobs AI can't replaceCement masons score lowest, at 8.
- Jobs most at risk from AIProofreaders score highest, at 76.
- High-paying jobs with low AI exposureRadiologists earn a median $420,860 and score 46.
- Entry-level jobs with low AI exposureCement masons need little prior training and score 8.
- Jobs AI speeds up the mostFor natural sciences managers, AI speeds up 96% of task time.
- AI and jobs in Malta15 iGaming roles graded task by task.
- Skills that keep work humanThe skills to build in the jobs AI changes least.
- Jobs A to ZEvery job we publish, by first letter.
- AI exposure by industryAverage scores for 22 occupation groups.
- Write a CV screening software can readHow applicant tracking systems read a CV, and what to change.
- AI for businessesThe tasks to automate first in each team and industry.
- How we score AI exposureThe grading, the weights and every source.