Skip to content
Radio Logo
PodcastsTechnologyThe Low Down

The Low Down

Low Level
Latest episode

20 episodes

  • The Low Down

    This SSH Bug Accepted Wrong Passwords, AI Exploits, Tiny Keys & a Surprise Guest

    17/11/2026 | 1h 10 mins.
    Recorded on Friday, November 13, 2026
    The Low Down Podcast is presented by Maze.
    https://mazehq.com/lowdown
    X: https://twitter.com/Maze_Security
    257 A's should not get you into an SSH account.
    John Hammond joins Ed and Matt to talk about N-able, ScreenConnect, and a rough week for remote support. Then we get into the SSH login bug, OpenAI's Linux kernel research, FortiClient, and eMudhra's tiny private keys. There's also some World of Warcraft, because apparently we needed a break.
    Today we're talking about:
    John Hammond, N-able, and ScreenConnect
    John walks through the RMM mess, including SCKit and the difference between a vulnerable server extension and a vulnerable technician client. We talk about how remote support can spread an infection, then get into incident-response burnout.
    N-able report: https://www.huntress.com/blog/n-able-disables-cloud-services-following-disclosure-of-able2shell-exploit-chains
    SCKit research: https://www.huntress.com/blog/sckit-worm-rogue-screenconnect
    Client advisory: https://www.connectwise.com/company/trust/security-bulletins/2026-09-08-screenconnect-bulletin
    Exploitation in the wild and local AI
    Why doesn't every scary bug turn into an internet-wide fire? We talk about exploit conditions, GreyNoise's PHP observations, a separate AI-assisted intrusion report, and running local models for security research.
    GreyNoise PHP tracking: https://php.events.greynoise.io/
    Operation Snowball: https://malbearlabs.com/posts/operation-snowball-inside-an-autonomous-ai-intrusion-campaign
    The SSH login bug
    Ed explains how OpenSSH and musl can turn a password longer than 256 bytes into a successful login on affected configurations. We follow the asterisks, the old fixes, and the practical problem of figuring out which devices are exposed.
    Original write-up and comic: https://sshhole.com/
    OpenAI's kernel bugs and Daybreak
    OpenAI reports 948 working local privilege escalation exploits. We unpack what that number means, the API-credit cost, and the research pipeline. Then we ask what happens to AI security vendors when the model companies start offering the tools themselves.
    Linux kernel research: https://openai.com/index/a-thousand-fewer-paths-to-root/
    Daybreak Red Team: https://openai.com/daybreak/red-team/
    Daybreak Defense Factory: https://openai.com/daybreak/defense-factory/
    FortiClient and the security-vendor problem
    A FortiClient/FortiPAM exploit demo leads into a bigger argument about edge devices, old toolchains, and why preventable bugs keep shipping in security products.
    Researcher's demo: https://www.linkedin.com/posts/kevin-joensen_cve-2026-121529-or-fortinet2shell-as-we-ugcPost-7526030522238697472-3JP7/
    eMudhra's weak private keys
    One private key was 4. We talk about the HSM migration mistake behind eMudhra's weak intermediate-certificate keys, what the hardware was supposed to protect, and why checking the result matters.
    Company incident report: https://bugzilla.mozilla.org/show_bug.cgi?id=2083449#c82
    World of Warcraft and logging off
    Alliance or Horde, busy-dad gaming, and finding time to play something without turning it into another job.
    Which is harder to keep up with right now: the vulnerabilities, the patches, or the AI product launches?
    Linked reports may include later updates.
    ----------
    0:00 Intro and John Hammond
    3:59 N-able and ScreenConnect attacks
    13:48 Exploitation in the wild
    17:28 Jailbroken AI and local models
    22:48 The 257-character SSH login bug
    32:42 OpenAI’s Linux kernel exploit research
    44:03 Daybreak and AI security vendors
    49:55 FortiClient and edge security
    58:14 eMudhra and the private key of 4
    1:04:13 World of Warcraft
    ----------
    Listen on Spotify!
    https://go.lowdownpod.com/spotify
    The Low Down is presented by Maze.
    LinkedIn: https://www.linkedin.com/company/mazehq/
    X: https://twitter.com/Maze_Security
    Follow Us!
    https://www.instagram.com/lowdown.pod
    Ed / Low Level
    Instagram: https://www.instagram.com/lowleveltv/
    LinkedIn: https://www.linkedin.com/in/lowleveltv/
    YouTube: https://www.youtube.com/@LowLevelTV
    Matt Johansen / Vulnerable U
    Instagram: https://www.instagram.com/mattjayy/
    LinkedIn: https://www.linkedin.com/in/matthewjohansen/
    YouTube: https://www.youtube.com/@VulnerableU
    John Hammond
    YouTube: https://www.youtube.com/@_JohnHammond
  • The Low Down

    Binance's Fake Interview Hack, Railway's AI Incident, Hacked Freezers & a Critical Linux Bug

    06/11/2026 | 1h 6 mins.
    The Low Down Podcast is presented by Maze.
    https://mazehq.com/lowdown
    X: https://twitter.com/Maze_Security
    The "signup test" was probing for Kubernetes secrets.
    Ed and Matt get into Railway's AI testing incident, hacked refrigeration systems, and fake job interviews targeting developers. Then Ed works through a Linux kernel bug before we finish with ARTEX and a warning about a prebuilt Docker image.
    Today we're talking about:
    Railway and the AI testing incident
    We follow the logs, the apologies, and the personal credit card that helped identify the employee. Josh Manders says every probing attempt was blocked and nothing was accessed; Railway later described it as an employee's personal project. Ed and Matt have plenty to say about authorization and the "my agent did it" explanation.
    Employee's account (https://x.com/jakerunzer/status/2117474337682780643)
    Josh's account (https://x.com/joshmanders/status/2117530391183507484)
    Railway's statement (https://x.com/JustJake/status/2117746864745578719)
    Credit-card exchange (https://x.com/vxunderground/status/2117656394396778956)
    Hacked freezers, blood banks, and exposed controls
    More refrigeration attacks lead into the practical problems of securing operational systems. We talk about slow patch approvals, unclear ownership, and what getting a management interface off the public internet actually involves.
    Supermarket report (https://www.bbc.com/news/articles/cq1kd1nee80eo)
    Blood-storage report (https://www.theguardian.com/technology/2026/nov/04/france-national-blood-service-hack-cyber-attack-europe)
    Binance and fake job interviews
    Binance's account of an employee compromised through a fake interview leads into how fake recruiters target developers. Matt walks through separate malicious coding-test and VS Code examples, including the pressure to run unfamiliar code while an interviewer watches. We discuss keeping those exercises off your everyday machine.
    Binance report (https://www.reuters.com/world/binance-says-hackers-used-fake-job-interview-compromise-computer-before-63-2026-11-03/)
    MetaPlay case study (https://github.com/alexhraber/metaplay-fake-interview-supply-chain-incident)
    VS Code case study (https://ransom-isac-landing.pages.dev/blog/contagious-interview-vscode-to-rat/)
    Linux TCP, MPTCP, and exploit math
    Ed explains a kernel use-after-free, the demonstrated MPTCP exploit, and why reliable exploitation is more complicated than the severity score suggests. We get into the conditions an attacker needs, patching, and some increasingly elaborate math.
    Go Security Team advisory (https://groups.google.com/g/golang-announce/c/PCoQcdxoScw/m/aD-MgXEvAAAJ)
    ARTEX and the Docker image warning
    CrowdStrike reports on ARTEX in real intrusions. Separately, researchers raise backdoor concerns about a specific prebuilt Docker image. We look at what they found and Jason Haddix's SpikeStrip experiment for tripping up AI pentesting tools.
    CrowdStrike research (https://www.crowdstrike.com/en-us/blog/vapor-panda-uses-artex-to-scale-global-campaign/)
    Image warning (https://x.com/GrahamHelton3/status/2117782531856581004)
    SpikeStrip announcement (https://x.com/Jhaddix/status/2117342447445627141)
    If you work with grocery-store or other operational networks, we'd like to hear what getting these systems off the internet looks like on the ground.
    Linked reports may include later updates.
    —————-
    Listen on Spotify!
    https://go.lowdownpod.com/spotify
    The Low Down is presented by Maze.
    LinkedIn: https://www.linkedin.com/company/mazehq/
    X: https://twitter.com/Maze_Security
    Follow Us!
    https://www.instagram.com/lowdown.pod
    Ed / Low Level
    Instagram: https://www.instagram.com/lowleveltv/
    LinkedIn: https://www.linkedin.com/in/lowleveltv/
    YouTube: https://www.youtube.com/@LowLevelTV
    Matt Johansen / Vulnerable U
    Instagram: https://www.instagram.com/mattjayy/
    LinkedIn: https://www.linkedin.com/in/matthewjohansen/
    YouTube: https://www.youtube.com/@VulnerableU
  • The Low Down

    Black October Recap, OpenCode’s Stolen API Keys, AI Distillation & the LiteLLM Supply-Chain Attack

    03/11/2026 | 1h 24 mins.
    Welcome to The Low Down, the best show on the internet for hackers
    Black October has real bugs, stolen keys, and a lot of very loud headlines.
    Ed and Matt sort through PHP2Shell, AI security claims, and the difference between a scary vulnerability and one attackers can actually use. Harry Wetherald, co-founder and CEO of Maze, joins us to talk about finding the bugs that matter.
    ——————————
    Today we're talking about:
    PHP2Shell and the Black October panic
    We get into working exploits, questionable proof-of-concepts, patching, and the reporting around attacks. Ed breaks down why configuration and exploit reliability matter, while Matt revisits a report he shared that was later corrected.
    Original research: https://calif.io/php2shell/
    Exploit validation: https://www.vulncheck.com/blog/php2shell-vulnerabilities
    Unit 42 report and updates: https://unit42.paloaltonetworks.com/cve-2026-109618-cve-
    026-109619/
    DREAM intrusion research: https://dreamgroup.com/blog/chained-php2shell-exploit-roots-25-servers
    AI security with Harry from Maze
    Harry joins us to talk about what AI can do for security teams, how to decide whether a vulnerability is exploitable in your environment, and why producing more findings isn't enough.
    Maze: https://mazehq.com/
    Brotli, Node.js, and GoAhead
    A decompression bug puts Brotli in the headlines, then GoAhead gives us a lesson in broken TLS. We talk through the technical details and the conditions that turn these bugs into real problems.
    Node.js security release: https://nodejs.org/en/blog/vulnerability/october-2026-security-releases
    GoAhead research: https://foxio.io/blog/choose-your-backend-three-tls-bugs-in-goahead
    OpenCode, stolen keys, and distillation
    OpenCode's Console incident leads to a discussion about stolen API keys and model distillation. Matt revisits his initial response and the distinction between stealing keys and what attackers intended to do with them afterward.
    OpenCode announcement: https://x.com/opencode/status/2111519637049319818
    Matt's clarification: https://x.com/mattjay/status/2111554937452990593
    LiteLLM, Railway, and the update problem
    More supply-chain trouble raises an uncomfortable question: how quickly should you update when new releases can carry the attack? We get into credential exposure, dependency hygiene, and the tradeoffs of waiting.
    LiteLLM attack-chain report: https://docs.litellm.ai/blog/security-incident-oct-21-2026-attack-chain
    Railway incident report: https://blog.railway.com/p/preliminary-report-october-security-incident
    WPS, pre-boot bugs, and AI bank access
    We cover Broadcom WPS research, an EDK II PXE vulnerability, and AI agents accessing bank accounts. The banking discussion turns on whose permission was given: the customer's or the bank's.
    WPSplinter research: https://qkaiser.github.io/broadcom-wps/
    EDK II advisory: https://github.com/tianocore/edk2/security/advisories/GHSA-99cj-7hg9-mp4m
    AI banking report: https://www.axios.com/2026/10/22/ai-agents-banks-unauthorized-accounts-access
    We finish with the noise around security news, incident-response fatigue, and what happens when every week is treated like the end of the internet.
    Recorded October 28, 2026.
    Linked reports may include later updates.
    Listen on Spotify!
    https://go.lowdownpod.com/spotify
    The Low Down is presented by Maze.
    LinkedIn: https://www.linkedin.com/company/mazehq/
    X: https://twitter.com/Maze_Security
    Follow Us!
    https://www.instagram.com/lowdown.pod
    ——————————
    Ed / Low Level
    Instagram: https://www.instagram.com/lowleveltv/
    LinkedIn: https://www.linkedin.com/in/lowleveltv/
    YouTube: https://www.youtube.com/@LowLevelTV
    Matt Johansen / Vulnerable U
    Instagram: https://www.instagram.com/mattjayy/
    LinkedIn: https://www.linkedin.com/in/matthewjohansen/
    YouTube: https://www.youtube.com/@VulnerableU
  • The Low Down

    Cybersecurity Experts REACT to the php chaos... (and MORE)

    28/10/2026 | 1h 27 mins.
    Welcome to The Low Down, the best show on the internet for hackers
    Listen on Spotify!
    https://go.lowdownpod.com/spotify
    The Low Down is presented by Maze.
    LinkedIn: https://www.linkedin.com/company/mazehq/
    X: https://twitter.com/Maze_Security
    ----------------
    Today we're talking about:
    PHP2Shell and the Disclosure Fight
    Ed and Matt are at OneCon in Las Vegas as PHP2Shell sets off a chaotic week. We unpack the research, shaky AI exploit claims, Chainguard's early fixes and whether disclosure timelines can keep up. Dan Lorenc and the PHP Foundation give different accounts of the rollout.
    https://calif.io/php2shell/
    https://www.linkedin.com/posts/danlorenc_i-wanted-to-share-what-happened-with-the-share-7518756991520120832-QX2_/
    https://thephp.foundation/blog/2026/10/21/todays-php-security-release-and-how-were-strengthening-php-security/
    Kyle Polley's embargo discussion: https://x.com/kpolley/status/2113069186318577666
    Your Certificate Authority's Private Key Was 4
    One of eMudhra's private keys was literally four. We talk tiny keys, certificate trust, the company's exit from SSL/TLS and what happens when clients still trust compromised certificates.
    https://scotthelme.co.uk/eight-bits-of-trust-emudhras-ca-keys-cracked-in-milliseconds/
    https://emudhra.we-have-a-problem.com/
    https://emudhra.com/en/press-release/emudhra-to-exit-global-ssl-tls-certificate-business
    https://www.mitmproxy.org/posts/releases/mitmproxy-12.3/
    Linux Bugs and the Power of a Pattern
    DirtyFrag, research credit and the latest kernel discussion lead into why Ed thinks AI is useful for spotting use-after-free patterns.
    DirtyFrag research: https://github.com/V4bel/dirtyfrag
    KSMBD background: https://docs.kernel.org/filesystems/smb/ksmbd.html
    Does AI Favor Attackers or Defenders?
    SentinelLABS' Tom Hegel makes the case for keeping powerful tools accessible to defenders. SentinelOne's Steve Stone explains why finding more bugs isn't the whole story: attack paths, speed, scale and the physics of an intrusion still matter.
    Meet the Threat Hunting World Champion
    Jason Walker joins us straight from his OneCon win. We get into his move from graphic design to security, CTF strategy, AI-assisted detection engineering and why the fundamentals still matter.
    OneCon: https://www.onecon.io/
    ----------------
    Follow Us!
    https://www.instagram.com/lowdown.pod
    Ed / Low Level
    Instagram: https://www.instagram.com/lowleveltv/
    LinkedIn: https://www.linkedin.com/in/lowleveltv/
    YouTube: https://www.youtube.com/@LowLevelTV
    Matt Johansen / Vulnerable U
    Instagram: https://www.instagram.com/mattjayy/
    LinkedIn: https://www.linkedin.com/in/matthewjohansen/
    YouTube: https://www.youtube.com/@VulnerableU
    Tom Hegel / SentinelLABS
    LinkedIn: https://www.linkedin.com/in/tom-hegel/
    Steve Stone / SentinelOne
    LinkedIn: https://www.linkedin.com/in/stevenstone618/
    Jason Walker / tresscross
    LinkedIn: https://www.linkedin.com/in/jasonwalker777/
    Blog: https://tresscross.blog/
  • The Low Down

    Cybersecurity Experts BREAK DOWN the shocking OpenAI Spying Incident, Crypto Backdoors, AND MORE

    19/10/2026 | 1h 16 mins.
    Welcome to The Low Down, the best show on the internet for hackers
    Listen on Spotify!
    https://go.lowdownpod.com/spotify
    The Low Down is presented by Maze.
    LinkedIn: https://www.linkedin.com/company/mazehq/
    X: https://twitter.com/Maze_Security
    Follow Us!
    https://www.instagram.com/lowdown.pod
    Heading to OneCon 2026? SentinelOne’s cybersecurity conference runs October 20–22 at the ARIA Resort & Casino in Las Vegas. Use code oc26lowdown for a discount on registration.
    https://www.onecon.io/
    Today we're talking about:
    AI Got Better While We Were Arguing About It
    Matt returns from SecTor, Ed explains his monitor-firmware experiment, and we discuss what AI can do for security research—and the human work it still takes.
    https://www.youtube.com/watch?v=wIe3eDfGKUo
    https://blog.cloudflare.com/cyber-frontier-models/
    https://hacks.mozilla.org/2026/05/behind-the-scenes-hardening-firefox/
    Your Hardware Wallet Has Extra Hardware
    Kim Zetter's reporting reveals seed-stealing Ledger implants, including devices from authorized retailers, and separate Trezor tampering from unauthorized sellers. Who planted them remains unknown.
    https://www.zetter-zeroday.com/who-is-planting-backdoors-in-ledger-crypto-wallets/
    https://www.zetter-zeroday.com/trezor-wallets-also-compromised-with-implants/
    Someone Hacked the Supermarket Fridges
    French supermarkets lose food after a cyberattack. We discuss exposed refrigeration controllers, a date-derived login code, and the separate Ahold Delhaize outage.
    Reuters report: https://srnnews.com/french-supermarket-chain-intermarche-hit-by-cyberattack-150-stories-targeted/
    https://claroty.com/team82/research/freeze-the-controller-defrost-the-food-uncovering-vulnerabilities-in-danfoss-refrigeration-controllers
    https://www.pentestpartners.com/security-blog/you-dont-need-a-nation-state-to-warm-up-a-fridge/
    https://www.bleepingcomputer.com/news/security/retail-giant-ahold-delhaize-takes-systems-offline-after-cyberattack/
    If AI Can Clone Your App, What Are You Selling?
    ArtCraft and Reverse Engineer Anything spark a debate about paid software, reliability and support. Fireship's FilmCraft test puts the hype under pressure.
    https://www.kdnuggets.com/the-rise-of-ai-software-clones-why-rea-and-artcraft-are-going-viral
    https://github.com/morluto/rea
    https://www.youtube.com/watch?v=wajoP_5Cpw8
    The People You Call After Ransomware
    We discuss Edward Dubrovsky's reported arrest and the separate MonsterCloud case. Prosecutors allege MonsterCloud secretly paid attackers while promising special decryption tools. These are allegations, not convictions.
    https://krebsonsecurity.com/2026/10/fbi-arrests-founder-of-ransomware-negotiation-firm/
    https://www.bleepingcomputer.com/news/security/ransomware-recovery-ceo-charged-over-secret-ransom-payments/
    https://www.justice.gov/opa/pr/known-cybersecurity-expert-and-owner-florida-ransomware-remediation-company-charged
    OpenAI Finds a Recorder Behind the TV
    An employee is accused of planting a recorder in an executive conference room. We unpack the alleged chatbot trail, a second device and the problem of insider access.
    https://www.bbc.com/news/articles/cwg5jq4vlddzo
    https://x.com/OpenAINewsroom/status/2110095403198849132
    A Jailbreak That Isn't Quite a Jailbreak
    hrtowii's iOS research involves Apple's Security Research Device signing system. The author explains the limits and says the bug has been patched.
    https://www.hrtowii.dev/blog/making-a-not-jailbreak-in-3-weeks
More Technology podcasts
About The Low Down
the internet's best podcast about hacking

Listen to The Low Down, Machine Gods and many other podcasts from around the world with the radio.net app

Get the free radio.net app

  • Stations and podcasts to bookmark
  • Stream via Wi-Fi or Bluetooth
  • Supports Carplay & Android Auto
  • Many other app features
The Low Down: Podcasts in Family