SilentGlass Unveiled: The NCSC’s First Hardware Warrior Blocks HDMI Malware in its Tracks + Video

Listen to this Post

Featured Image

Introduction:

The HDMI cable on your desk does more than just transmit video – it also carries a hidden, bidirectional 100 Mbit/s Ethernet channel (HDMI Ethernet Channel, or HEC) along with the Consumer Electronics Control (CEC) protocol. These “invisible” data paths, often overlooked by security teams, create a genuine attack surface: threat actors can exploit them to inject malicious commands, exfiltrate data via compromised displays, or pivot into internal networks through a seemingly benign monitor. To counter this, the UK’s National Cyber Security Centre (NCSC) has developed SilentGlass, a plug‑and‑play inline device that actively inspects and blocks unexpected or malicious traffic on HDMI and DisplayPort connections, finally treating physical display links as a true security boundary.

Learning Objectives:

  • Understand the hidden security risks of HDMI’s Ethernet (HEC) and control (CEC) channels, and how they can be weaponised by attackers
  • Learn to deploy and verify inline HDMI security devices like SilentGlass, including practical commands to monitor EDID/CEC activity on Linux and Windows
  • Explore TEMPEST side‑channel attacks that recover on‑screen text from electromagnetic emissions, and implement countermeasures using deep‑learning defences

You Should Know:

  1. The Hidden Threat: HEC, CEC, and Why Your Monitor is a Security Boundary

The HDMI specification includes several features beyond video transmission. The HDMI Ethernet Channel (HEC) creates a bidirectional 100 Mbit/s (or 1 Gbit/s in newer versions) network link between a source and a display. Alongside it, the Consumer Electronics Control (CEC) protocol allows devices to control each other – for example, turning on the TV when a PC is powered up. Both channels operate at the hardware level, often invisible to firewalls and endpoint detection software.

Attackers can exploit these channels in several ways:

  • Malicious EDID payloads – An attacker‑controlled display can send malformed EDID (Extended Display Identification Data) to trigger buffer overflows in GPU drivers. Several CVEs exist for Linux kernel vulnerabilities triggered by corrupted EDID data.
  • CEC command injection – A compromised monitor can generate CEC remote‑control keystrokes, which some hosts translate into actual keyboard inputs, potentially allowing an attacker to execute commands.
  • HEC network pivot – If both ends support HEC, an attacker with access to the display could establish a direct Ethernet link to the host machine, bypassing network‑level controls.

Mitigation techniques

  • Disable unnecessary HDMI features – In UEFI/BIOS, turn off CEC (often labelled “HDMI‑CEC” or “Control”) and any “HDMI Ethernet” options. On Windows, navigate to Device Manager → Monitors → Properties → Driver and disable any “CEC Device” or “HEC Adapter” if present.
  • Use physical inline filters – Devices like SilentGlass enforce policy at the hardware level, blocking any unexpected traffic regardless of software state.
  • Monitor EDID/CEC traffic with open‑source tools – The NCC Group’s CECster tool (Python‑based, uses a USB‑CEC bridge) allows you to fuzz and inspect HEC/CEC communications. Install it with:
 Linux / macOS / Windows (with Python 3)
git clone https://github.com/nccgroup/CECster.git
cd CECster
pip install -r requirements.txt
python CECSTeR.py

This GUI tool can send arbitrary CEC frames, replay EDID blocks, and detect anomalous responses, helping you audit the security of connected displays.

  1. Step‑by‑Step: Deploying and Verifying an Inline HDMI Security Device

The following guide assumes you have a physical inline HDMI monitor (e.g., SilentGlass) placed between a computer and a display. In its default “block” mode, it allows only standard video/audio traffic and denies any HEC or CEC packets. You can test its effectiveness using a simple CEC injection script on Linux.

Linux – Checking CEC device presence and sending a test command

1. Install cec-utils (available on most distributions):

sudo apt install cec-utils  Debian/Ubuntu
sudo dnf install cec-utils  Fedora

2. List CEC adapters – Connect your host to the display via the inline device and run:

cec-client -l

This will show detected CEC devices (e.g., /dev/cec0). If the inline device is functioning correctly, you should not see the connected monitor – the CEC traffic is being blocked.
3. Attempt to send a “power on” command (if the inline device allowed CEC):

echo "on 0" | cec-client -s -d 1

With a properly configured inline blocker, the display should not respond, confirming that CEC packets are dropped.
4. Monitor raw EDID – Use `edid-decode` to inspect the EDID that the host sees:

sudo apt install edid-decode
cat /sys/class/drm/card0-HDMI-A-1/edid | edid-decode

The inline device may present a “clean” EDID, blocking any malicious extensions from the real display.

Windows – PowerShell approach

  1. Check for connected monitors and their EDIDs remotely (requires admin privileges):
    Get-WmiObject -Namespace root\wmi -Class WmiMonitorID | Select-Object 
    
  2. More detailed EDID inspection using the `Get‑EDID.ps1` script from the Microsoft Device Console tools:
    Save and run the script
    .\Get-EDID.ps1
    
  3. Test CEC / HEC behaviour – Because Windows does not expose CEC natively, use the Python `pyCEC` library:
    pip install pyCEC
    

Then run a simple test script:

import cec
cec.init()
cec.tx(cec.CEC_CMD_IMAGE_VIEW_ON)

A well‑functioning inline device will prevent any response from the display.

  1. Deep‑TEMPEST: How Attackers “See” Your Screen Through the Air

While SilentGlass blocks malicious traffic originating from a display, another class of attacks – TEMPEST – eavesdrops on the electromagnetic radiation leaking from HDMI cables themselves. In 2024, researchers at the Universidad de la República showed that a Software Defined Radio (SDR) and a deep‑learning model can reconstruct on‑screen text by capturing leaked signals from an HDMI cable several meters away. Their open‑source implementation, Deep‑TEMPEST, improves text character error rates by over 60 percentage points compared to previous methods.

How it works (simplified):

  • An attacker places a low‑cost SDR (e.g., HackRF One or RTL‑SDR) near the target’s HDMI cable.
  • The SDR captures electromagnetic emanations in a specific frequency band (e.g., 300‑500 MHz for HDMI).
  • A pre‑trained neural network maps the captured signal back to the original pixel values, recovering readable text or images.

Practical countermeasures

  • Physical shielding – Use high‑quality shielded HDMI cables, and ensure that computer cases and monitor chassis are properly grounded. For high‑security environments, install ferrite chokes on all video cables.
  • Distance and attenuation – TEMPEST attacks typically require the antenna to be within a few metres. Maintaining physical security perimeters reduces risk.
  • Signal obfuscation – Tools like `gr‑tempest` (a GNU Radio flowgraph) can be used to test your own emissions. Install it on Linux:
    git clone https://github.com/git-artes/gr-tempest.git
    cd gr-tempest
    mkdir build && cd build
    cmake .. && make && sudo make install
    

    Then run the `automatic_tempest_example.grc` flowgraph in GNU Radio Companion to analyse your environment’s leakage.

What Undercode Say:

  • Hardware trust boundaries matter – The NCSC’s SilentGlass legitimises the idea that physical video connections must be treated as a point of control, not an assumed trust boundary. In industrial and government environments, this is a paradigm shift.
  • Defence in depth includes the physical layer – While software patching and network firewalls are essential, they cannot block a malicious EDID that crashes a GPU driver. Inline hardware enforcements close a gap that software alone cannot fill.
  • TEMPEST is no longer just a spy‑agency problem – With SDRs costing under $300 and open‑source deep‑learning models available, electromagnetic eavesdropping is now in the reach of sophisticated attackers. Organisations handling sensitive data must reassess physical security for displays.
  • Expect more “hardware cybersecurity” products – SilentGlass is the first consumer‑grade device from the NCSC’s IP licensing programme, but it won’t be the last. We anticipate similar inline filters for USB, Thunderbolt, and other multi‑protocol cables.

Prediction:

The commercialisation of SilentGlass signals a broader industry trend: hardware‑enforced security policies will become standard for any interface that carries both data and control signals. Expect to see “secure cables” or “smart dongles” for USB‑C, DisplayPort, and Thunderbolt that apply zero‑trust principles at the physical layer. At the same time, the democratisation of TEMPEST tools will force organisations to adopt AI‑driven signal obfuscation and improved electromagnetic shielding as part of their standard security posture. Over the next 3‑5 years, treating every cable as a potential attack vector will move from a niche government concern to a mainstream cybersecurity requirement.

▶️ Related Video (82% Match):

🎯Let’s Practice For Free:

IT/Security Reporter URL:

Reported By: Rob Hulsebos – Hackers Feeds
Extra Hub: Undercode MoN
Basic Verification: Pass ✅

🔐JOIN OUR CYBER WORLD [ CVE News • HackMonitor • UndercodeNews ]

💬 Whatsapp | 💬 Telegram

📢 Follow UndercodeTesting & Stay Tuned:

𝕏 formerly Twitter 🐦 | @ Threads | 🔗 Linkedin | 🦋BlueSky