← Back to Feed CACHED · 2026-07-01 06:42:49 · CACHE_KEY CVE-2026-20459
CVE-2026-20459 · CWE-288 · Disclosed 2026-07-01

In Modem

ASSESSED — NOISGATE V0.5
Vendor
Reassessed
Verdict:
Do you agree?
01 · The Real Story

You need to be a fake cell tower to crash one phone's radio for as long as it stays in range

This is a MediaTek baseband bug (Patch ID MOLY01816800, Issue MSV-6842) in the modem's handling of malformed over-the-air signaling. A UE that camps on an attacker-controlled base station and receives a crafted message crashes the modem — the phone drops the cellular link until it reboots or re-attaches. Affected products span MediaTek's mobile chipset families shipped in Android handsets and IoT/CPE modems; MediaTek's July 2026 security bulletin is the canonical list. CWE is classified as 288 (auth bypass via alternate path) in NVD, but the practical primitive is improper input validation leading to DoS, not code execution or persistent auth compromise.

There is no vendor CVSS baseline for this CVE, so noisgate is establishing the first assessment. The impact ceiling is denial of service on a single handset within RF range of a rogue eNB/gNB — no RCE, no privilege escalation, no data disclosure documented, no lateral movement, no persistence past re-attach. Real-world severity is LOW: the exploit prerequisite (a rogue base station within physical proximity of a targeted device) is the same barrier that gates every Stingray-class attack, and the payoff is only a reboot.

"Modem DoS via rogue base station — attacker needs a Stingray to knock one phone offline. Backlog hygiene, not an emergency."
02 · The Attack Path

4 steps from start to impact.

STEP 01

Stand up a rogue base station

Attacker deploys a fake eNodeB/gNodeB or 2G/3G downgrade cell using srsRAN, OpenAirInterface, or a commercial IMSI-catcher (Stingray, Harris StingRay II). This requires an SDR (USRP B210/X310, LimeSDR, BladeRF) plus RF amplifiers strong enough to out-signal the legitimate carrier at the target's location. Regulatory: transmitting in licensed bands is a felony in most jurisdictions.
Conditions required:
  • SDR + amplifier hardware
  • software radio stack (srsRAN/OAI)
  • physical proximity to target
  • willingness to violate spectrum regulations
Where this breaks in practice:
  • hardware cost $2k-$50k
  • detection by carrier RF monitoring and Android's NetworkNotifyType.CELLULAR_SUSPICIOUS_NETWORK
  • target device may prefer 5G-SA and refuse to camp on a rogue 4G/LTE cell
Detection/coverage: CellularPrivacy AIMSICD, SnoopSnitch, Android 12+ null-cipher warning, carrier-side rogue-cell detection
STEP 02

Coerce target UE to camp

Rogue cell advertises stronger RSRP/RSRQ than surrounding legitimate cells so the target's baseband selects it during idle-mode reselection or handover. For iron-fisted coercion, jam the legitimate carrier's downlink while radiating the rogue cell.
Conditions required:
  • target UE within RF range
  • rogue cell signal dominant
  • target not locked to 5G-SA-only mode
Where this breaks in practice:
  • modern UEs pin to home PLMN when possible
  • corporate MDM can disable 2G/3G fallback
  • target moves out of range
Detection/coverage: SIM/PLMN change events surfaced by MDM and by Android's Advanced Protection cell-security controls
STEP 03

Send crafted signaling message

Attacker transmits the malformed NAS/RRC frame that trips the input-validation flaw (MOLY01816800). Public PoC does not exist; researcher would need to fuzz the MediaTek baseband — tooling analogs include LTEFuzz, BaseSpec, and FirmWire.
Conditions required:
  • knowledge of the exact malformed field(s)
  • ability to encode and transmit the frame post-attach
Where this breaks in practice:
  • baseband internals are closed-source
  • no public PoC or exploit primitive released
Detection/coverage: essentially none from the defender's side — baseband is a black box
STEP 04

Modem crash — DoS

Baseband panics, the cellular radio drops, and the UE loses voice/SMS/data until modem restart or full device reboot. Impact ends when the target moves away from the rogue cell or reboots; no persistence, no code execution, no data disclosure documented.
Conditions required:
  • successful delivery of malformed frame
Where this breaks in practice:
  • automatic modem reset on many MediaTek platforms restores service in seconds
Detection/coverage: user-visible loss of signal; MDM/EMM cellular telemetry
03 · Intelligence Metadata

The supporting signals.

In-the-wild exploitationNone observed. No campaigns, no threat-actor attribution, no incident reports as of 2026-07-01.
Public PoCNone. No GitHub repo, no Metasploit module, no writeup. Baseband fuzzers (LTEFuzz, BaseSpec, FirmWire) could theoretically reproduce.
EPSSNot yet scored (freshly disclosed 2026-07-01). Expect <1st percentile — DoS + physical-proximity precondition.
KEV statusNot listed. Unlikely to be added — no RCE, no privilege escalation, no ransomware relevance.
CVSSNo vendor score. noisgate assesses ~3.8 (AV:A/AC:H/PR:N/UI:N/S:U/C:N/I:N/A:L) — adjacent RF, high complexity, availability-only, single-device scope.
AffectedMediaTek modem software across chipset families listed in MediaTek's July 2026 Product Security Bulletin. Patch ID MOLY01816800, Issue MSV-6842.
Fixed versionsMediaTek MOLY baseband builds carrying MOLY01816800. OEM firmware SPLs from Samsung/Xiaomi/Oppo/Vivo dated 2026-07-05 or later.
Exposure telemetryNot internet-reachable — RF-only vector. Shodan/Censys/GreyNoise irrelevant. Exposure = installed base of vulnerable chipsets on hosts physically reachable by a rogue cell.
Disclosed2026-07-01 in MediaTek's monthly bulletin, coordinated with Android Security Bulletin cycle.
ReporterCredited to MediaTek internal security research (per bulletin format); no external researcher named.
04 · The Call

noisgate verdict.

Final Verdict
= UNCHANGED to LOW (3.8/10)

The single decisive factor is the physical-proximity precondition: exploitation requires a rogue base station within RF range of the target device, and the impact ceiling is a single-device denial of service with no code execution or data exposure. That combination floors the verdict at LOW — this is a Stingray-class attack whose payoff is only a phone reboot.

HIGH impact ceiling is DoS only
HIGH attack vector requires RF proximity
MEDIUM affected chipset list (MediaTek bulletin still being enumerated by OEMs)

Why this verdict

  • No vendor baseline — establishing first assessment. Neither NVD nor MediaTek published a CVSS score; noisgate scores ~3.8 based on the documented primitive.
  • Attack vector is Adjacent (RF), not Network. Exploitation requires a rogue base station within antenna range of the victim UE — this is Stingray-class tradecraft, not internet-reachable.
  • Impact is availability-only, single device. No RCE, no privilege escalation, no data disclosure, no persistence past modem reset. Vendor description explicitly says *system crash* / *remote DoS*.
  • Role multiplier: consumer handset (typical role). Chain succeeds → one phone loses signal until reboot. Blast radius: local, single-device. No fleet impact.
  • Role multiplier: IoT/CPE/M2M modem in a fleet (high-value role). Chain succeeds → attacker with an RF footprint over a warehouse or industrial site could brown-out cellular-connected sensors within that footprint. Blast radius still bounded to RF coverage; no lateral movement, no data compromise. Does not cross the HIGH floor — no domain, identity, hypervisor, or supply-chain surface is touched.
  • No KEV, no PoC, no EPSS traction. Zero indicators of imminent operational risk.

Why not higher?

MEDIUM/HIGH would require either an internet-reachable vector, code execution, or persistent impact — none are present. The rogue-base-station prerequisite alone caps this well below MEDIUM for any threat model that isn't a targeted-surveillance operation against a specific individual.

Why not lower?

IGNORE would require the bug to be non-exploitable in practice or the affected footprint to be negligible. MediaTek chipsets ship in hundreds of millions of Android handsets and IoT devices, and journalists, executives, and dissidents genuinely are targeted by IMSI-catchers — a documented DoS primitive is a real (if narrow) tool in that playbook.

05 · Compensating Control

What to do — in priority order.

  1. Apply July 2026 SPL to managed Android fleet on normal cadence — The MediaTek fix rides the Android Security Bulletin (SPL 2026-07-05). LOW verdict carries no noisgate mitigation SLA — go straight to the 365-day remediation window and let normal MDM patch rings handle it.
  2. Disable 2G/3G fallback on high-risk-user handsets — For executives, journalists, or personnel travelling to high-surveillance regions, enforce 5G-SA-preferred and 2G-off via MDM (Android Enterprise setCellularTwoGDisabled, iOS Lockdown-equivalents). Removes the easiest downgrade path an IMSI-catcher uses to force camp. Deploy alongside the next MDM policy push.
  3. Enroll high-risk users in Android Advanced Protection or equivalent — Surfaces rogue-cell / null-cipher warnings and blocks unencrypted cellular where available. Deploy for VIP cohort now — no reason to wait on this regardless of CVE.
  4. Inventory MediaTek chipsets in IoT/CPE/M2M fleet — Router, POS, and industrial-IoT deployments often use MediaTek basebands and lag Android SPL. Confirm carrier/OEM patch availability so these devices are not stranded when the vendor patch ships.
What doesn't work
  • VPN / MDM tunnel enforcement — the crash happens at Layer 1/2 of the cellular stack before any IP traffic exists. VPN cannot intercept baseband signaling.
  • Network firewall / IDS / WAF — none of these touch the RF air interface.
  • MFA / phishing training — attack requires no user interaction and no credential.
  • Airplane mode as a mitigation — technically prevents attach, but also disables the device's primary function. Not a real control.
06 · Verification

Crowdsourced verification payload.

Run this on an auditor workstation with Android Debug Bridge (adb) in PATH and a target Android device connected via USB with USB debugging enabled. Invoke as .\Check-CVE-2026-20459.ps1 -SerialNumber <adb-device-id>. No elevated privileges required; the script reads ro.build.version.security_patch and the modem baseband version.

noisgate-verify.ps1
POWERSHELLREAD-ONLYSAFE
#!/usr/bin/env pwsh
# Check-CVE-2026-20459.ps1
# Detects whether an attached Android device carries the July 2026 SPL that
# includes MediaTek Patch ID MOLY01816800 (CVE-2026-20459).
#
# Exit codes: 0 = PATCHED, 1 = VULNERABLE, 2 = UNKNOWN

param(
    [Parameter(Mandatory=$true)]
    [string]$SerialNumber
)

$ErrorActionPreference = 'Stop'

function Invoke-Adb {
    param([string]$Cmd)
    return (& adb -s $SerialNumber shell $Cmd 2>$null).Trim()
}

try {
    $spl      = Invoke-Adb 'getprop ro.build.version.security_patch'
    $baseband = Invoke-Adb 'getprop gsm.version.baseband'
    $chipset  = Invoke-Adb 'getprop ro.hardware'
    $vendor   = Invoke-Adb 'getprop ro.vendor.mediatek.platform'
} catch {
    Write-Output 'UNKNOWN: adb query failed'
    exit 2
}

Write-Output "Device      : $SerialNumber"
Write-Output "Chipset     : $chipset"
Write-Output "MTK platform: $vendor"
Write-Output "Baseband    : $baseband"
Write-Output "SPL         : $spl"

# Only MediaTek-based devices are in scope
$isMtk = ($chipset -match 'mt[0-9]{4}') -or ($vendor -ne '')
if (-not $isMtk) {
    Write-Output 'PATCHED (not MediaTek — CVE-2026-20459 not applicable)'
    exit 0
}

# Fix rides Android Security Bulletin 2026-07-05
try {
    $splDate = [datetime]::ParseExact($spl, 'yyyy-MM-dd', $null)
} catch {
    Write-Output 'UNKNOWN: could not parse SPL'
    exit 2
}

$fixed = [datetime]'2026-07-05'
if ($splDate -ge $fixed) {
    Write-Output 'PATCHED'
    exit 0
} else {
    Write-Output 'VULNERABLE — SPL predates 2026-07-05; MOLY01816800 not present'
    exit 1
}
07 · Bottom Line

If you remember one thing.

TL;DR
Verdict is LOW (3.8) and there is no vendor baseline — this is the first assessment. Per the noisgate mitigation SLA for LOW there is no mitigation deadline; treat this as backlog hygiene. Per the noisgate remediation SLA, patch within 365 days via normal Android Security Patch Level cadence (target SPL 2026-07-05 or later on MediaTek-based handsets). Monday morning: (1) confirm your MDM patch rings will pick up the July 2026 Android bulletin on the usual schedule, (2) inventory MediaTek chipsets across your Android + IoT/CPE fleet so you know who is actually in scope, and (3) for your VIP / high-surveillance-risk user cohort, push a 2G/3G-disable MDM policy now — that policy costs nothing and removes the most common IMSI-catcher downgrade primitive independent of this CVE. Do not burn a change window on this.

Sources

  1. OffSeq Threat Radar — CVE-2026-20459
  2. MediaTek Product Security Bulletin (July 2026)
  3. Android Security Bulletin — July 2026
  4. NVD — CVE-2026-20459
  5. MITRE CVE Record — CVE-2026-20459
  6. CWE-288 — Authentication Bypass Using an Alternate Path or Channel
  7. srsRAN Project (rogue-cell tooling context)
  8. BaseSpec / FirmWire baseband fuzzing research
Peer Review

What defenders are saying.

Submit a review attribution: handle + country only
0 flags selected · stored anonymously
Validation Results

Crowdsourced verification outputs.

Results submitted by users who ran the verification payload against their environment.