XHack Logo
XHack
Products
Services
Compliance
Pricing
Resources
Company
Sign upLogin
XHack Logo
XHackOffensive Security

Certified offensive security team delivering penetration testing evidence written for your auditor.

OSCP+OSCPC-AI/MLPenCASA
support@xhack.io

24/7 SOC Operations

XHack Status
Under attack? Get help now
Services
  • VAPT Services
  • Red Teaming
  • SOC Services
  • Threat Intelligence
  • Incident Response
  • Managed Testing
Pricing
  • Platform Plans
  • Services Pricing
Compliance
  • SOC 2
  • PCI DSS
  • ISO 27001
  • GDPR
  • HIPAA
  • ISO 42001
  • AI Maturity Assessment
  • TX-RAMP
  • NBFC / SECP
  • All Frameworks
Products
  • Vulnerability Assessment
  • GitGuard
  • AI Probe
  • SOC Dashboard
  • AI Agent
  • Cloud Investigation
Comparison
  • XBOW vs XHack
  • Horizon3 vs XHack
  • Strix vs XHack
  • Pentera vs XHack
Resources
  • Platform Tour
  • All Features
  • Install the Agent
  • XHack AI
  • Documentation
  • Blog
  • Case Studies
  • Documents
  • FAQ
Company
  • About Us
  • Our Team
  • Certifications
  • Security and Trust
  • VAPT Explained
  • Contact

© 2026 XHACK (SMC-PRIVATE) LIMITED. All rights reserved.

Registered with the SECP, CUIN 0361945. Verify

Security & TrustVulnerability DisclosurePrivacy PolicyTerms of ServiceRefund Policy
Blog/Security

Offensive AI: What the 2026 Evidence Shows and How to Use It Legally

XHack

XHack

Author

November 18, 2026

15 min read

Offensive AI: What the 2026 Evidence Shows and How to Use It Legally

Table of contents

15

What Offensive AI Means

What Attackers Are Doing With Offensive AI

How Good Is Offensive AI Really? Three Sources, Three Caveats

Where Offensive AI Still Fails

Using Offensive AI Legally

What Defenders Should Change Because of Offensive AI

How XHack Approaches Offensive AI

FAQ: Offensive AI Questions Answered

What is offensive AI?

Is offensive AI legal?

How dangerous is offensive AI in 2026?

Can offensive AI replace penetration testers?

What should defenders do about offensive AI?

Where can I use offensive AI legally?

The Bottom Line

By Salman Khan, OSCP+, Founder of XHack, SRT (Synack Red Team member)

Read this in 30 seconds: Offensive AI means using AI on the attack side of security, and it has two very different users: criminals who misuse models, and authorised testers who use them to find weaknesses first. The 2026 evidence says the technical gap is closing fast, but the thing that separates a dangerous operator from a hobbyist is no longer skill. It is orchestration. This guide covers what the data shows, where offensive AI still fails, how to use it legally, and what defenders should change.

  • Misuse is rising, and it is mostly early-stage. Anthropic’s June 2026 analysis of 832 banned accounts found the share rated medium risk or higher rose from 33% to 56% across the study window, while lateral movement appeared in only 6.5% of accounts.
  • Capability is climbing. Irregular reports Cybench success rising from 10% in early 2024 to 82% in November 2025, but says models still struggle with long-horizon planning, noisy environments and error recovery.
  • Human-level on one network, not on every network. In the Stanford ARTEMIS study, one agent configuration beat 9 of 10 professional testers, with 82% of its findings valid, on a single university network.
  • The legal line has not moved. Offensive AI is lawful for systems you own or are authorised to test, and a crime against anyone else’s. Verified routes exist, and jailbreaking is not one of them.
  • Defenders need a faster clock. If attackers move faster, patch the exploited list first, and a short script below shows how to check your CVEs against it.

The scary part is not that it hacks like a genius, it is that it lets one operator do the work of a team.

Search for this topic and you get two kinds of page: breathless warnings and thin tool lists. Neither helps if you are deciding what to defend, what to test with, or what you are allowed to do. This guide sticks to primary sources and says plainly where the numbers come from companies that sell something.

It contains no attack techniques and no jailbreaks. I run XHack, so I also say where we fit and where I would trust the evidence less.

Offensive AI capability animation: Cybench success rising from 10% in early 2024 to 82% in November 2025, and the hard private task tier rising from near zero to about 60% by late fall 2025
Offensive AI capability on public and private tests, as reported by Irregular, 2024 to late 2025

What Offensive AI Means

Offensive AI is the use of AI to perform or speed up attack-side security work, such as reconnaissance, exploit development, social engineering or post-compromise activity. The same technology serves two groups, and mixing them up is the source of most confusion.

  • Authorised offensive security. Penetration testers, red teams and bug bounty researchers using AI on systems they own or have written permission to test, to find weaknesses before criminals do.
  • Misuse. Criminals and state-linked groups using commercial or open-weight models to build tools, evade detection and process stolen data, which providers try to detect and ban.
Offensive AI two meanings compared: authorised testing against misuse, who does it, what is allowed, and how providers and law treat each
Offensive AI has two users: authorised testers and attackers, and the law treats them very differently

The technology is the same and the legality is not. That is why the rest of this guide keeps returning to authorisation.

What Attackers Are Doing With Offensive AI

The best public dataset on misuse comes from Anthropic, and it shows attackers using AI mostly to build and hide tooling, not to run whole intrusions. In a 3 June 2026 post, Anthropic’s Frontier Red Team analysed 832 accounts it banned for cyber-related policy violations between March 2025 and March 2026, mapping 13,873 observed actions to MITRE ATT&CK.

The findings, as Anthropic reports them:

  • Early stages dominate. Developing capabilities, mostly malware development, appeared in 69% of accounts, obfuscation in 64.7%, and defence evasion in 84.4%.
  • Late stages are rare. Lateral movement appeared in 6.5% of accounts, and impact and exfiltration were each under 3% of observations.
  • Risk is rising. The share of accounts rated medium risk or higher rose from 33% to 56% between the first and second halves of the window, about 1.7 times.
  • Skill is not the divider. Anthropic’s conclusion is that the line between low-risk and high-risk actors is “no longer technical skill but orchestration.”

Three caveats apply. These are only accounts Anthropic banned, so they show one provider’s view. The rise may partly reflect better detection, which Anthropic acknowledges. And the risk score is Anthropic’s own composite, so treat it as a lens, not a measurement of attack success.

Offensive AI misuse animation: early-stage techniques such as malware development and defence evasion dominate banned accounts while lateral movement is rare, with medium-risk share rising from 33% to 56%
What 832 banned accounts did with AI, per Anthropic’s June 2026 analysis: heavy on tooling, light on late-stage intrusion

How Good Is Offensive AI Really? Three Sources, Three Caveats

Offensive AI is clearly improving, and every source that says so has a reason to be read carefully.

Irregular reports that success on the public Cybench benchmark rose from 10% in early 2024 to 82% in November 2025, and that on its own private “hard” task tier success went from near zero to roughly 60% by late fall 2025. Irregular is an evaluation company, its page names no authors or model lists, and its date metadata conflicts with its body, so I treat the direction as credible and the exact figures as indicative.

The Stanford ARTEMIS study (arXiv 2512.09882) is the strongest independent-style test. On a live network of about 8,000 hosts, one ARTEMIS configuration scored second overall, ahead of 9 of 10 professional testers, with 82% of its findings valid. Another configuration had 55% valid, and agents were more prone to false positives than humans. It is one network, a small sample and the authors’ own scoring.

XBOW’s mid-year 2026 report, as covered by CyberScoop, says the miss rate on its web exploitation benchmark fell from 40% for GPT-5 to 10% for GPT-5.5. XBOW sells offensive security tooling and I did not open the report itself, so this is a vendor claim reported second-hand.

Put together: offensive AI is strong at well-defined tasks, and the trend is steep.

Where Offensive AI Still Fails

The same sources agree on the limits, and they matter more than the headlines. Irregular says models still struggle in open-ended or noisy environments, with long-horizon planning, state management and recovering from errors, and that performance drops as attack chains grow. Irregular, citing Anthropic’s account of a large AI-orchestrated campaign that Anthropic scored as maximum risk, notes the model hallucinated vulnerabilities and misdescribed target data. ARTEMIS had a higher false-positive rate than humans and struggled with graphical interfaces.

That gives defenders something to work with. It is fast and cheap per attempt, and it is unreliable without a human or a validation step. The sensible reading is not “AI hacks everything”, it is “an attacker with AI makes more attempts, faster, with the same blind spots”.

Test it on your own target

The XHack agent finds, verifies and reports real vulnerabilities in web apps, APIs, hosts and mobile.

Start free for 7 days

Using Offensive AI Legally

Offensive AI is legal for authorised work and illegal against anyone else, and the tool changes nothing about that. In most countries unauthorised access to a computer system is a crime, for example under the US Computer Fraud and Abuse Act and the UK Computer Misuse Act. This is not legal advice, so check your local law and contract.

Four habits keep you on the right side:

  1. Written scope first. A signed agreement or program scope that names the targets.
  2. Stay inside it. An unlisted host is not yours to test, and nor is a third-party service you depend on.
  3. Use verified access for sensitive work. Providers now offer it, and it asks who you are.
  4. Keep records and a stop button. Know how to halt the run, and keep logs.

The verified routes are real. OpenAI’s Daybreak program offers reduced refusals for verified defensive work (Blue) and authorised offensive testing (Red), though its help page says individuals can apply for Blue only, with a paid plan, Advanced Account Security, a FIDO2 key and identity verification, and approval is not automatic. Anthropic’s June post describes routing higher-risk dual-use activity through a Cyber Verification Program so defenders can keep working. Our guide on why AI refuses hacking requests covers the rules in more depth.

What Defenders Should Change Because of Offensive AI

If offensive AI lowers the cost of an attempt, shrink the window an attacker can use. Five changes follow directly from the evidence above.

  • Patch the exploited list first. CISA’s Known Exploited Vulnerabilities catalog is the clearest signal of what is being used right now, and recent entries in our coverage carried due dates of only three days, such as CVE-2026-97229 in GitLab and CVE-2026-119977 in WordPress Core.
  • Cut exposed management interfaces. Cheap, fast attempts hit whatever is reachable.
  • Protect AI provider keys and sessions. Treat them like cloud credentials, because anyone holding them can run models and spend money as you.
  • Watch for orchestration, not just techniques. A noisy, automated sequence of low-skill steps can still add up to an intrusion.
  • Test yourself with offensive AI on authorised scope, so you see what an attacker’s tooling would see.
Five changes defenders should make because of offensive AI: patch the exploited list first, cut exposed management interfaces, protect AI keys, watch for orchestration and test yourself
Five defender changes that follow from the offensive AI evidence

Here is a short script for the first item. It downloads CISA’s public KEV file and checks a list of your CVEs against it, printing the due date and whether the entry carries the forensic-triage flag. It is read-only, and I ran it against live data on 13 November 2026.

import json
import sys
import urllib.request
from datetime import date

URL = "https://www.cisa.gov/sites/default/files/feeds/known_exploited_vulnerabilities.json"


def load():
    req = urllib.request.Request(URL, headers={"User-Agent": "Mozilla/5.0 kev-check"})
    with urllib.request.urlopen(req, timeout=60) as r:
        return json.load(r)


def main():
    args = sys.argv[1:]
    if args and args[0] == "--file":
        args = [l.strip() for l in open(args[1]) if l.strip() and not l.startswith("#")]
    wanted = [a.upper() for a in args]
    data = load()
    kev = {v["cveID"]: v for v in data["vulnerabilities"]}
    today = date.today()
    print(f"catalog {data['catalogVersion']}, {data['count']} entries, checked {today}")
    print(f"{'cve':16} {'added':11} {'due':11} {'status':10} forensic  product")
    for cve in sorted(wanted, key=lambda c: kev.get(c, {}).get("dueDate", "9999")):
        v = kev.get(cve)
        if not v:
            print(f"{cve:16} {'-':11} {'-':11} {'not listed':10} -         (not in KEV, which is not the same as safe)")
            continue
        due = date.fromisoformat(v["dueDate"])
        status = "PAST DUE" if due < today else f"{(due - today).days}d left"
        print(f"{cve:16} {v['dateAdded']:11} {v['dueDate']:11} {status:10} {v.get('forensicTriage', '-'):9} {v['vendorProject']} {v['product']}"[:140])


if __name__ == "__main__":
    main()

Run it with your own CVE IDs:

python3 kev_check.py CVE-2026-116606 CVE-2026-111560 CVE-2026-119977 CVE-2026-97229 CVE-2026-111357 CVE-2026-12345
catalog 2026.11.10, 1779 entries, checked 2026-11-13
cve              added       due         status     forensic  product
CVE-2026-97229   2026-10-29  2026-11-01  PAST DUE   Yes       GitLab Community Edition and Enterprise Edition
CVE-2026-111357  2026-10-30  2026-11-02  PAST DUE   Yes       Ivanti Connect Secure
CVE-2026-119977  2026-11-03  2026-11-06  PAST DUE   Yes       WordPress WordPress Core
CVE-2026-116606  2026-11-10  2026-11-13  0d left    Yes       Barracuda Networks Barracuda Email Security Gateway (ESG)
CVE-2026-111560  2026-11-10  2026-11-13  0d left    Yes       Barracuda Networks Multiple Products
CVE-2026-12345   -           -           not listed -         (not in KEV, which is not the same as safe)
KEV check animation: a terminal looks up five CVEs in CISA's Known Exploited Vulnerabilities catalog and flags three past due and two due today
Checking a CVE list against CISA’s catalog: three entries past due, two due today, one not listed

“Past due” here means the federal due date has passed, which is a useful urgency signal for anyone, not a legal deadline for you. And “not listed” never means safe.

How XHack Approaches Offensive AI

I am biased, it is my company, and I think XHack is the strongest way to use offensive AI legally, because verification is built in and a person stands behind the result. XHack combines three things most providers sell separately: an autonomous AI agent, a continuous security platform and human testers.

Here is what I can state from our public pages.

  • Verified access. The agent finds vulnerabilities from the start, and exploitation unlocks once your identity is verified, with one 7-day trial per person. Unrestricted means verified and logged, not anonymous.
  • The agent runs the work. It plans and acts with up to 12 sub-agents by default and three browsers, and XHack does not store your agent chats.
  • Humans when it counts. Managed testing by OSCP-certified researchers starts at $2,500, with patch verification and compliance-ready reporting included. Essential covers one unauthenticated web app or up to 50 host IPs in 4 days.
  • Plans. Individual plans are $20, $49 and $150 a month, and company plans start at $560 a month.

If you want to see what an attacker with offensive AI would see on your estate, run it on a scoped target you own first. See our AI vs human penetration testing guide, the pricing page and our AI VAPT services. Brutal honesty is kind of our thing.

FAQ: Offensive AI Questions Answered

What is offensive AI?

Offensive AI is the use of AI for attack-side security work such as reconnaissance, exploit development and post-compromise activity. It is used by authorised testers to find weaknesses, and by criminals who misuse models. The technology is the same, and the legality depends entirely on authorisation.

Is offensive AI legal?

Using it on systems you own or have written permission to test is legal, and using it against anyone else’s systems is a crime in most countries. This is not legal advice, so check your local law and your contract.

How dangerous is offensive AI in 2026?

It is meaningfully more capable than in 2024, but it is not autonomous magic. Anthropic’s data shows misuse is mostly early-stage tooling, Irregular reports models still struggle with long-horizon planning and noisy environments, and agents produce false positives. The danger is scale and speed, not perfect skill.

Can offensive AI replace penetration testers?

Not on current evidence. In the Stanford ARTEMIS study one agent configuration beat 9 of 10 testers on one network, with more false positives than humans and weak results on graphical tasks. Offensive AI is best for breadth and repetition, with a person validating results and owning scope.

What should defenders do about offensive AI?

Shrink the attacker’s window. Patch CISA’s Known Exploited Vulnerabilities first, remove exposed management interfaces, protect AI provider keys and sessions, and test your own systems with authorised offensive AI so you see what an attacker would see.

Where can I use offensive AI legally?

On your own systems, in a written scope you have agreed, in a bug bounty program within its rules, in CTFs and labs, and through verified platforms and provider programs such as OpenAI’s Daybreak or XHack’s identity-verified agent. Never test a system without permission.

The Bottom Line

Offensive AI is not one thing. It is a fast, cheap assistant in the hands of both testers and attackers, strongest at well-defined tasks and weakest at long, messy operations. The evidence says orchestration, not skill, now separates the dangerous from the dabbling, so defenders should shrink the window and testers should work inside a written scope. Use offensive AI where you are authorised, through routes that verify who you are, and treat every result as a lead until a human has confirmed it.

Put this into practice

Point the XHack agent at a target you are authorised to test. It hunts, proves and writes the report.

Start free for 7 days

Categories

Security

Previous

Jailbroken AI for Security Research: What It Costs You and the Legal Route

Next

NodeZero vs Pentera: Pricing, Scope and Who Each One Fits in 2026

On this page

What Offensive AI Means

What Attackers Are Doing With Offensive AI

How Good Is Offensive AI Really? Three Sources, Three Caveats

Where Offensive AI Still Fails

Using Offensive AI Legally

What Defenders Should Change Because of Offensive AI

How XHack Approaches Offensive AI

FAQ: Offensive AI Questions Answered

What is offensive AI?

Is offensive AI legal?

How dangerous is offensive AI in 2026?

Can offensive AI replace penetration testers?

What should defenders do about offensive AI?

Where can I use offensive AI legally?

The Bottom Line

Related articles

Continue reading

AI Red Team Operations: Where AI Fits in a Real Engagement and Where It Does Not

Security

AI Red Team Operations: Where AI Fits in a Real Engagement and Where It Does Not

AI red team operations explained: which of 6 phases AI speeds up, what a 40-network study found, DORA rules for testers ...

Read article
Jailbroken AI for Security Research: What It Costs You and the Legal Route

Security

Jailbroken AI for Security Research: What It Costs You and the Legal Route

Jailbroken AI for security research: the five costs, what research shows on defences and the verified route to fewer ref...

Read article
NodeZero vs Pentera: Pricing, Scope and Who Each One Fits in 2026

Security

NodeZero vs Pentera: Pricing, Scope and Who Each One Fits in 2026

NodeZero vs Pentera compared from vendor pages and AWS listings: published per-asset prices, scope, certifications, revi...

Read article