Episodes

  • The FBI Warned About ARTEX. AI Is Collecting Our Security Debt
    Nov 19 2026

    "ARTEX is exactly what I described on Fox Business last month when I called AI a security debt collector."

    It is cashing in on decades of vulnerabilities, misconfigurations and poor hygiene that we as an industry never got around to fixing.

    On November 13th, the FBI put out a FLASH alert on ARTEX, an open source AI penetration testing tool. AWS and Cloudflare published their own data the same day. Two days earlier, Palo Alto's Unit 42 published its own analysis.

    In this Resilient Cyber solo I walk through what ARTEX actually is, what the major threat intel teams are seeing, the real-world impact, why ARTEX will have company, and how the fundamentals, done exceptionally well, plus AI and deception, can help defenders fight back.


    Key takeaways


    ARTEX is a harness that ships with no model of its own. You give it a goal in plain language, planner agents break that goal into tasks, and worker agents run ordinary tools like Nmap, curl and a headless browser. The operator plugs in whatever model they want.


    On October 8th the developer pulled the original repository. Cloudflare calls what happened next a hydra effect, with over 12,000 repositories now in the ecosystem, and says ARTEX and its forks make up over 60 percent of the AI pentest orchestrators it knows about.


    The FBI says some intrusions went from initial targeting to privileged access in under three hours, a figure that comes from Microsoft. In a separate case, Microsoft watched an agent go from searching GitHub for client secrets to pulling data out of Azure Key Vault in five minutes.


    The impact landed in about four months. Seoul police tied a seven-person gang led by a former pentester to more than 90 organizations and 8.2 million records. CrowdStrike tracked a China-nexus group it calls VAPOR PANDA running a single ARTEX controller against more than 200 organizations in at least 14 countries over 11 days. ARTEX is one name among many. Unit 42 analyzed 14 tools in the ecosystem and in October counted 609 internet-facing hosts running ARTEX and 1,323 running CyberStrike AI. Cloudflare is tracking more than 50 of these AI agents.


    Every one of those reports says the techniques are old. Exposed portals, hardcoded keys, missing rate limits and unpatched edge systems. Qualys says its customers fix critical vulnerabilities in a median of seven days, which is still 168 hours against an attacker who needs a few. The practitioner answer is the fundamentals, done exceptionally well, plus AI and deception on defense. In Tracebit's October research, nine misdirection techniques cut completed objectives by 91 percent and multiplied the attacker's model spend per objective by 23.


    Chapters


    0:00 Four reports, one week

    0:47 What ARTEX actually is

    2:09 What threat intel teams are seeing

    3:56 The real-world impact

    5:16 ARTEX will have company

    6:24 The debt collector shows up

    7:48 The fundamentals, done exceptionally well

    9:28 Fighting back with AI and deception

    10:53 Where this leaves us


    Links


    Substack


    https://www.resilientcyber.io/p/from-industrialized-discovery-to-operations


    YouTube


    https://www.youtube.com/watch?v=jzHrvsqQa7Q


    Subscribe to Resilient Cyber


    https://www.resilientcyber.io


    FBI FLASH alert


    https://www.ic3.gov/CSA/2026/261113.pdf

    Show more Show less
    12 mins
  • Zero Trusty for AI Agents Starts at the Endpoint | Rob Allen
    Nov 16 2026

    Ask a ThreatLocker customer whether OpenClaw is running on their machines and the answer takes one sentence: if you didn't explicitly allow it, it isn't. Rob Allen on why default deny may be the only sane answer to agentic AI and a flood of vulnerabilities.

    Rob joined ThreatLocker in 2021, when it was about 25 people. It's now more than 750. In that time, the customer question has flipped from "keep these agents off our machines" to "the business wants Claude Code, Codex, and Antigravity, so make them safe."

    We get into why you can't patch your way out of this year, how ring fencing shut down remote code execution through PHP2Shell, and what happens when an AI agent quietly spins up a VM your host can't see into.

    In this episode:

    ● Script kiddies feeding PHP2Shell into AI for exploits, and why most of what came out was slop

    ● The friend in Ireland who asked Rob to install OpenClaw, and why he flatly refused

    ● Why a vendor questionnaire is still just a questionnaire until someone verifies it

    ● Locking your company's Claude tenancy to one or two IP addresses so a stolen token is useless

    ● Whack-a-mole on steroids: why waiting for patches is a losing strategy
    ● Assume every piece of software you run is full of holes, and act accordingly
    ● 1,400 fixes in a single Patch Tuesday, and broken patches that make teams afraid to update
    ● Ring fencing: letting PHP run but stopping it from calling anything else
    ● Monitor-only mode, so locking things down doesn't break the business
    ● The MCP query that revealed a customer approving 900 of 1,000 requests by switching off protection and granting admin rights
    ● The WSL exfiltration trick, and why Claude Cowork's sandbox VMs raise the same problem
    ● A Claude process tree that looks like a forest, and one policy that controls Claude plus everything it spawn
    ● Why users click yes, yes, yes on every agent permission prompt

    Chapters:

    0:00 Intro
    01:55 From 25 people to 750 at ThreatLocker
    1:08 AI cuts both ways, and the PHP2Shell slop
    4:36 Vendor reviews in the age of agents
    7:38 Security versus utility
    10:40 How ring fencing stopped PHP2Shell
    16:39 Metrics ways of reporting
    20:30 ThreatLocker's AI control release
    21:43 Agent sandboxes and the WSL exfil trick
    24:40 One policy for Claude and everything it calls

    Connect with Rob Allen:

    ThreatLocker: https://www.threatlocker.com

    Resilient Cyber: https://www.resilientcyber.io

    Subscribe for more conversations with security practitioners and leaders.

    #zerotrust #aisecurity #agenticai #endpointsecurity #applicationallowlisting #ciso

    Show more Show less
    27 mins
  • What Cyber Founders Get Wrong About Marketing | Gianna Whitver
    Nov 9 2026

    Gianna Whitver, co-founder of the Cybersecurity Marketing Society, on why technical founders can't fully hand off their company's story, and why your first marketing hire won't magically fix the pipeline.

    I sit down with Gianna Whitver, who runs the Cybersecurity Marketing Society. She describes it as an ISAC for marketers. It started as a Slack channel with about 10 friends she met at RSA and Black Hat, after she realized the B2B tech marketing that worked in other industries didn't work in cyber. It has since grown to more than 5,000 members, almost entirely through word of mouth.

    We get into what founders need to own in go-to-market, how to talk about AI to skeptical security buyers without it sounding like hand-waving, and how marketing should be measured at each stage of a company.

    In this episode:

    ● Why the founder's underdog story and vision have to come from the founder, even when the founder isn't a natural storyteller
    ● Communicating AI to security buyers through proof points and outcomes rather than buzzwords
    ● When to make the first marketing hire, and why seed stage is earlier than many founders think
    ● Why expecting a marketer to fix the pipeline immediately misses that the message or the product may be the real problem
    ● The mismatch of hiring a late-stage CMO into an early-stage startup
    ● How much time founders should put into their personal brand, and how that changes from $5M to $20M ARR
    ● Using AI for content without losing an original point of view
    ● How marketing KPIs should shift from target account penetration early on to revenue later

    Chapters:

    0:00 Intro
    1:40 From a Slack channel to 5,000+ members
    3:42 The story only the founder can tell
    6:00 What gets a skeptical buyer's attention
    8:49 When to make the first marketing hire
    10:20 Common mistakes with early marketing hires
    13:00 How much founders should invest in their personal brand
    16:04 The founder's role from $5M to $20M ARR
    16:58 Using AI for content without sounding like AI
    18:31 How founders should measure marketing
    19:42 CyberMarketingCon and the Cyber CEO Summit

    Connect with Gianna:

    LinkedIn: https://www.linkedin.com/in/giannawhitver
    CyberMarketingCon: https://cybermarketingconference.com

    Resilient Cyber: https://www.resilientcyber.io

    Subscribe for more conversations with security practitioners and leaders.

    #cybersecurity #gotomarket #startups #marketing #founders #cybersecuritymarketing

    Show more Show less
    23 mins
  • Is the Vulnpocalypse Real? Jeremiah Grossman Checks the Data
    Nov 2 2026

    Jeremiah Grossman, founder of Root Evidence, on why the vulnpocalypse hasn't shown up in the exploitation data, and why the security industry still has no efficacy data for nearly anything it sells.

    I sit down with Jeremiah Grossman, who has spent almost 30 years in cybersecurity. He started at Yahoo, built WhiteHat Security, which helped give birth to the AppSec industry, worked at SentinelOne, and sold an attack surface management company to Tenable. His latest company, Root Evidence, goes after vulnerability management because that is where the attackers have moved.

    We dig into his research showing CVE volume climbing while exploitation stays flat. He makes the case for building security around CVE data from cyber insurance claims tied to actual financial loss, and argues that the industry runs on consensus rather than evidence.

    In this episode:

    ● Why only about 1.48% of roughly 250,000 CVEs published since 2018 have been exploited
    ● Why attackers only needing 1 to 2% of CVEs to stay the number one vector is a giant fail for vulnerability management
    ● The roughly 600 CVEs tied to financial loss in insurance claims, and why those should be fixed first
    ● The defender window: a median of 116 days from patch to exploitation in 2026, up from 24 days in 2018
    ● What cyber insurers know about which controls work, and why they won't tell customers where they're wasting money
    ● Product warranties, and how a $5 million loss guarantee changes the patching conversation with the business
    ● Why "identity theft" shifts victimhood to the wrong party, and why there is still no catastrophe database for cyber
    ● What data would change his mind about AI-driven exploitation

    Chapters:

    0:00 Intro
    1:45 Jeremiah's background
    2:05 Why vulnerability management, and why now
    5:53 Attackers only need 1 to 2% of CVEs
    9:48 The defender window and why attackers are slowing down
    13:09 Cyber insurers and the missing efficacy data
    22:35 What will it take to wake the industry up?
    24:26 What would change his mind on the Vulnpocalypse
    26:22 Unprompted, SANS AI Security Summit, and what's next

    Connect with Jeremiah:

    LinkedIn: https://www.linkedin.com/in/grossmanjeremiah

    Root Evidence: https://rootevidence.com

    Resilient Cyber: https://www.resilientcyber.io

    Subscribe for more conversations with security practitioners and leaders.

    #vulnerabilitymanagement #cyberinsurance #aisecurity #riskmanagement #ciso #cybersecurity

    Show more Show less
    28 mins
  • The AI Cyber Apocalypse That Isn’t Coming | Ciaran Martin
    Oct 26 2026

    Ciaran Martin built the UK's National Cyber Security Centre from the ground up, led it for its first four years, and now teaches at the University of Oxford. Lately he's become one of the sharpest critics of the panic around frontier AI and cybersecurity. On this episode of the Resilient Cyber Show, I put his argument to the test.

    His case is that two things are true at once. Frontier AI is a real shift for attackers, and the reaction to it is the biggest FUD outbreak in over a decade.

    Look at the timeline. When the delayed release of Mythos was announced on April 7th, the warning was that defenses would be overrun within six months. Six months later, there has been no discernible rise in disruptive attacks, and disruptive attacks can't hide. Britain's worst damage still came from traditional attacks in 2025 on Jaguar Land Rover, Marks & Spencer, and the Co-op.

    Meanwhile, the hype is drowning out real incidents. The same week AI agents reading a few Australian websites made global headlines, the FBI appeared to have been breached by ShinyHunters, exposing case assignments, spousal addresses, and psychiatric evaluations. It barely registered.

    We get into why the famous 27-year-old bug wasn't worth $20,000 in tokens to any real attacker, and what OpenAI's engineers admitted about Hugging Face at Black Hat. We also cover the "cult of the machine god," alignment with whose values, and why software makers have dodged liability for decades. Ciaran also asks the question the AI debate keeps skipping: who would actually launch the catastrophe?

    Then he takes on the hardest question himself. What if he's wrong? His answer draws on running the NCSC through WannaCry and NotPetya in 2017. He closes with the one question every CISO should ask the board before discussing any "AI hacked X" headline.

    ⏱️ CHAPTERS

    00:00 Introduction
    01:37 Introducing Ciaran Martin
    02:02 The Early Days of Cybersecurity
    03:04 Why Governments Lagged on Cybersecurity
    04:55 Public-Private Cybersecurity Partnerships
    06:00 Why Leaders Avoid Technical Questions
    07:27 AI Hype vs. Cybersecurity Reality
    11:22 AI Safety vs. Cybersecurity
    19:04 Two ways of reading the AI
    25:46 The FBI Breach and AI Hype
    27:43 ShinyHunters and Real-World Cyber Threats
    29:37 Who Will Use AI to Launch Cyberattacks?
    29:58 Who Would Actually Launch the Catastrophe?
    35:00 The Devil's Advocate Case: AI-Powered Script Kiddies
    43:33 What CISOs Should Tell the Board
    45:00 Closure

    SOURCES AND FURTHER READING

    Ciaran Martin in The Economist — Fears of AI-induced armageddon are overdone: https://www.economist.com/by-invitation/2026/08/23/fears-of-ai-induced-armageddon-are-overdone

    📬 SUBSCRIBE TO RESILIENT CYBER
    https://www.resilientcyber.io

    If this was useful, subscribe. I cover AI security, agentic AI, AppSec, software supply chain, vulnerability management, and cyber policy for practitioners every week.

    #AISecurity #Cybersecurity #FrontierAI #CyberPolicy #AISafety #NCSC #ThreatIntelligence #CISO

    Show more Show less
    46 mins
  • AI Least Privilege: We Never Got It Right for Humans | Itamar Apelblat
    Oct 19 2026

    Itamar Apelblat was called back to reserve duty with Unit 8200 and spent hours waiting on access approvals. Then he remembered a service account his old team had set up seven years earlier. The password was a funny one, which is why he still knew it. He typed it in and became an admin on one of the most secure networks in the world.

    Nobody had changed that password. They hadn't forgotten about it. They were scared to touch it, because no one knew what would break.

    Now put that same fear in front of AI agents. When Token Security's research team checked its enterprise customers, more than 16 percent already had Muse running locally against corporate email and other resources. Muse launched three weeks before we recorded and is already the most popular productivity app in the App Store. Itamar expects agents to outnumber humans in the enterprise a hundred to one.

    His argument is that you can't secure agents by predicting every way they'll go wrong. An agent chases its goal like it's possessed and takes any access path that gets it there. What you can control is how much autonomy it has and what it's allowed to touch. Both come down to identity.

    In this episode I sit down with Itamar, co-founder and CEO of Token Security, and we get specific:

    → The three things companies find the first time they look: agents wired straight into production on raw API keys, abandoned experiment agents still holding privileges, and agents still running after their owner left the company
    → What intent-based access management actually means, and why an agent built to send a weekly report should never be able to delete from your database
    → Why IGA and PAM fall apart here: they treat Okta or Active Directory as the source of truth, while agents live on laptops, in SaaS, and in the cloud
    → What happens when a coding agent tries to connect to a production Snowflake environment, and how to decide where a human still has to approve
    → Who should own agent security, and why Itamar puts most of it with the IAM team

    I also put the hard question to him. We've had twenty years to get least privilege right for humans and mostly failed, so why would agents go any better? Then, at the end, he gives his advice on where security teams should start. It isn't the usual "get visibility first," and it's the part I'd most want a security leader to hear.

    ⏱️ CHAPTERS

    0:00 Introduction
    1:49 Unit 8200 and "Which Product Do You Hate the Most?"
    3:58 The Seven-Year-Old Password That Still Worked
    6:32 Autonomy and Access: The Two Dimensions of Agent Risk
    8:19 Why Agents Act Like They're Possessed
    10:14 16% Already Run Muse: What Companies Find
    12:46 Intent-Based Access Management, Explained
    15:15 Twenty Years of Failed Least Privilege
    17:15 Why Your IGA Won't Scale to Agents
    19:04 Where the Human Stays in the Loop
    21:12 Who Should Own Agent Security
    23:33 The Security Teams That Get This Right
    26:35 Closing Thoughts

    SOURCES AND FURTHER READING

    Token Security: https://www.token.security/
    Itamar Apelblat on LinkedIn: https://www.linkedin.com/in/itamar-apelblat/

    📬 SUBSCRIBE TO RESILIENT CYBER
    https://www.resilientcyber.io

    If this was useful, subscribe. I cover AI security, agentic AI, AppSec, software supply chain, vulnerability management, and cyber policy for practitioners every week.

    #AISecurity #AgenticAI #IdentitySecurity #NonHumanIdentities #IAM #AIAgents #LeastPrivilege #Cybersecurity

    Show more Show less
    27 mins
  • Your Annual Tabletop Exercise Is Already Outdated | Matt Hartley
    Oct 12 2026

    Matt Hartley, co-founder and Chief Product Officer of BreachRx, on why the once-a-year tabletop no longer matches the incidents organizations will actually face, and how AI-accelerated micro tabletops let teams train like they fight.

    In this episode I sit down with Matt Hartley, co-founder and Chief Product Officer of BreachRx. Matt has spent 28 years in cybersecurity. He started in government offensive security, then moved to the cyber threat intel startup iSight Partners, where he ran product through its acquisition by FireEye, and later ran Mandiant Engineering. He watched consulting teams parachute into companies that were compromised and meaningfully unprepared, and that convinced him the real gap was the response process, not detection or tooling.

    We get into what a micro tabletop actually looks like and why exercises should run in the same system you will use during a real incident. We also cover how notification clocks as short as 30 minutes change what legal teams need to practice, and why the floor is shifting toward multiple concurrent incidents.

    In this episode:
    ● Why the industry over-invests in technology and under-invests in the processes and procedures that carry a response
    ● What a micro tabletop looks like, from a one-hour cross-functional session to an exercise that runs asynchronously over a week with a 30-minute hot wash
    ● Train like you fight, and why rehearsing in your actual response platform lowers stress and folds lessons learned back into procedures
    ● More than 200 laws with incident notification requirements, including China's 30-minute window and India's six-hour rule covering even denial of service
    ● Starting a legal tabletop partway into a ransomware scenario to pressure test DORA, SEC, and other regulatory timelines
    ● Why privileging everything means nothing is privileged, and separating the factual record from privileged conversations
    ● The risk of 10 to 100 Slack or Teams channels spinning up around a single incident, and how one grouchy message can end up in evidence
    ● Keeping exercises from becoming check-the-box by matching frequency to team size and threat, pulling scenarios from the headlines, and running two incidents at once
    ● Why AI is raising the attack floor, from smarter background malicious traffic to lower-tier nation states making big capability jumps
    ● A workforce of AI agents that cuts incident reporting from 20 hours to 30 minutes and turns NIST, CISA, or Gartner guidance into tailored procedures
    ● A crawl, walk, run playbook for your first micro tabletop: three injects, one ornery stakeholder relationship, and a ten-minute debrief
    ● What regulators and plaintiffs will ask for, and why the SEC may look at five years of incidents, not just one

    Chapters:
    0:00 – Intro
    1:42 – Matt's Path from Offensive Security to Mandiant
    2:14 – Why Response, Not Detection, Was the Real Problem
    3:40 – What a Micro Tabletop Actually Looks Like
    7:37 – Train Like You Fight in the Real Platform
    9:33 – Regulatory Clocks, DORA, SEC, and 30-Minute Notifications
    11:56 – Privilege, Discoverability, and the 100 Slack Channel Problem
    15:03 – Keeping Exercises from Becoming Check-the-Box
    17:48 – Concurrent Incidents and a Rising Attack Floor
    19:53 – Where AI Agents Fit in Incident Response
    22:01 – Running Your First Micro Tabletop Next Month
    24:12 – Proving Continuous Readiness to Regulators and Plaintiffs
    26:55 – Closing

    Connect with Matt Hartley:
    LinkedIn: https://www.linkedin.com/in/wmhartl/

    BreachRx: https://www.breachrx.com
    BreachRx IR Exercises: https://www.breachrx.com/rex-platform/incident-response-exercises/

    Read my piece on BreachRx and incident response, Chaos, Clocks, and

    Command Centers: https://www.resilientcyber.io/p/chaos-clocks-and-command-centers

    Resilient Cyber: https://www.resilientcyber.io

    Subscribe for more conversations with security practitioners and leaders.
    #incidentresponse #tabletopexercise #aisecurity #ciso #cyberresilience #securityleadership

    Show more Show less
    27 mins
  • Who Is Winning the AI Security Market Right Now? James Berthoty
    Oct 5 2026

    James Berthoty of Latio on why the endpoint is now the center of AI security, and why most of the real agent risk traces back to blind spots we already had.

    I sit down with James Berthoty, who started Latio as an engineer frustrated by how hard it was to figure out what security tools actually did. That turned into a practitioner-focused, buyer-oriented analyst firm. We dig into Latio's second AI security market report. Last year the story was browsers and proxies. This year it's the endpoint, where coding agents now run with the permissions and credentials on developer laptops.

    We get into what roughly 400 AI security startups are actually differentiating on and why runtime detection claims need hands-on testing. We also cover why James pushes back on the runtime-first narrative, where the new AI security budget is coming from, and how security leaders should narrow the vendor field before they ever take a demo.

    In this episode:
    ● Why we are in the "pre-HTTPS era" of AI security, and how coding agents turned developer endpoints into the new pressure point
    ● Dedicated AI security budget jumping from 8% to 37% in a year, and how it flows out of AI governance
    ● What is already commoditized, and why intent-based runtime detection needs hands-on testing
    ● Agent blast radius, and why your agent shouldn't be able to drop the prod DB because you shouldn't be able to either
    ● Why recent AI safety incidents look more like misconfigurations than emergent behavior
    ● The next M&A wave as a fight over the laptop
    ● Anthropic's inference hooks, and why frontier labs likely won't absorb the security market
    ● MCPs and skills as a software supply chain problem with a new delivery mechanism

    Chapters:
    0:00 Intro
    0:31 Why James started Latio
    1:00 From browsers to the endpoint
    8:15 Do you need a purpose-built AI security platform?
    11:52 Where the AI security budget comes from
    13:53 400 startups and testing runtime claims
    19:28 Blast radius and runtime vs. posture
    23:35 M&A and the fight over the endpoint
    33:22 Inference hooks and frontier labs vs. vendors
    36:03 MCPs, skills, and software supply chain
    39:30 Advice for security leaders

    Connect with James:
    LinkedIn: https://www.linkedin.com/in/james-berthoty
    Latio: https://www.latio.com/
    AI Security Market Report: https://pulse.latio.tech/p/the-2026-latio-ai-security-market

    Resilient Cyber: https://www.resilientcyber.io

    Subscribe for more conversations with security practitioners and leaders.
    #aisecurity #agenticai #endpointsecurity #appsec #supplychainsecurity #mcp #cybersecurity

    Show more Show less
    41 mins