This is a conference-room remote left on every desk, not a master key to the building
CVE-2026-30496 is an unauthenticated control surface on the Optoma CinemaX P2 exposed over HTTP on TCP/2345. On confirmed vulnerable firmware TVOS-04.24.010.04.01 running Android 8.0.0, any host on the same local network can read settings and issue write operations across 74 endpoints, including power, volume, mute, brightness, display modes, and protocol toggles such as TelnetOn; the reporting researcher also notes likely exposure in sibling models sharing the same firmware platform, including CinemaX P1 / UHZ65UST and CinemaX Pro.
The vendor-style 9.8/CRITICAL framing does not match how this lands in real enterprise environments. The biggest friction is attacker position: this is same-network control of a projector, not unauthenticated internet-reachable server takeover, and the observed impact is device control/disruption rather than domain-wide compromise; that pushes this down to a MEDIUM operational-security issue unless your AV gear sits on flat user VLANs or hostile guest networks.
4 steps from start to impact.
Find the projector on an internal segment
nmap or simple subnet sweeps to identify hosts exposing TCP/2345 on an office LAN, guest Wi-Fi, or AV VLAN. The research write-up says the issue was found during an internal nmap sweep, which matches the real attacker workflow for this bug.- Attacker has access to the same Layer 2/LAN segment or a routed path to the projector
- The projector is powered on and network-connected over Wi-Fi or RJ45
- Most enterprises do not expose conference-room projectors directly to the public internet
- Segmentation between guest, user, and AV networks can block discovery entirely
- These devices are not high-prevalence compute assets compared with VPNs, firewalls, or hypervisors
2345/tcp is more reliable than CVE plugin coverage.Verify unauthenticated read access
curl, the attacker requests an endpoint like /get/Volume or /get/ProjectorID and receives data without any login challenge. The endpoint inventory shows the URL pattern is deterministic and lacks per-action authorization.- TCP/2345 is reachable
- The API process is listening normally
- Some NAC, east-west firewall, or wireless client isolation deployments will prevent direct peer access
- If the device is offline from the network, the bug is dead
Issue state-changing commands
curl -X PUT against /set/<action>?value=X or action endpoints to change settings without authentication. The researcher demonstrated write access to Volume, Mute, and protocol flags such as TelnetOn, and states every enumerated action reachable via read paths is also writable.- Knowledge of valid endpoint names
- Ability to send HTTP methods to the projector
- Impact is constrained to the projector rather than a broader identity or server control plane
- This is noisy: users notice power, audio, or display changes immediately
PUT /set/ requests to port 2345 in proxy, firewall, Zeek, or NetFlow-plus-packet telemetry; there is little reason for ordinary workstation traffic to hammer projector APIs.Disrupt meetings or weaken adjacent controls
- The room device is in active use or trusted by staff
- Operators do not immediately isolate the device
- Blast radius is typically one room or one projector at a time
- No public evidence currently shows mass exploitation campaigns or chained takeover beyond device control
The supporting signals.
| In-the-wild status | No public exploitation evidence found in the reviewed sources, and not listed in CISA KEV. |
|---|---|
| PoC availability | Yes, trivially reproducible: the disclosure includes working curl examples for unauthenticated read/write plus a full endpoint inventory. |
| EPSS | 0.0006 from the user-provided intel block, which is very low predicted near-term exploitation probability. |
| KEV status | Not KEV-listed as of the reviewed CISA catalog page; no CVE-2026-30496 entry located. |
| CVSS vector reality check | Published/adopted vector is CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:H, but the description repeatedly says any device on the same network. In practice this behaves closer to an adjacent/internal network exposure, not an internet-facing AV:N service. |
| Affected versions | Confirmed: CinemaX P2 firmware TVOS-04.24.010.04.01 on Android 8.0.0. Likely affected by shared firmware per researcher: CinemaX P1 / UHZ65UST and CinemaX Pro. |
| Fixed version | No vendor fix identified for CVE-2026-30496 in the reviewed material. The March 19, 2026 C13.2 release notes address HDMI/audio/app issues and the researcher states it does not close port 2345 control access. |
| Exposure and scanning | Everything points to internal LAN / AV network exposure, not broad internet exposure. I found no concrete Shodan/Censys/GreyNoise evidence in the reviewed sources showing meaningful internet-scale scanning or mass exposure. |
| Disclosure timeline | Researcher reports vendor notification on 2026-02-01, public disclosure on 2026-05-02, NVD publication on 2026-05-07. |
| Reporter | Stefan / whitelabel.org disclosed the issue and published both the advisory and endpoint inventory. |
noisgate verdict.
The decisive downshift is attacker position: exploitation requires reachability from the same local network or a routed internal path to the projector, which means this is usually a post-entry or insider abuse problem rather than an initial-access internet crisis. Impact is real but narrow—one AV device or room at a time—so the vendor's CRITICAL label overstates enterprise blast radius.
Why this verdict
- Major downward pressure: requires same-network access. The advisory and NVD text both describe control by any device on the same network, which implies internal foothold, shared Wi-Fi, guest access, or flat AV segmentation—not the classic internet-exposed unauthenticated server case implied by a 9.8 label.
- Blast radius is usually one device, one room. Even full control here mostly means nuisance, outage, or local AV sabotage; it is not demonstrated domain compromise, server takeover, or tenant-wide data loss.
- Threat telemetry is weak. EPSS is extremely low, KEV is negative, and I found no reviewed evidence of active campaigns or meaningful internet-scale scanning tied to this CVE.
- But it is still not LOW. No authentication is required, the API is deterministic and easy to script with
curl, and there is no identified vendor fix, so exposed projectors on user-accessible networks will absolutely get abused by bored insiders or post-compromise operators.
Why not higher?
There is no solid evidence of widespread external exposure, active exploitation, or reliable escalation from projector control to broader enterprise compromise. The requirement for local-network reachability is the compounding friction that breaks the vendor's CRITICAL story.
Why not lower?
This is not harmless information leakage. The API allows unauthenticated state changes across dozens of functions, including power, audio, image, and protocol settings, and the lack of a vendor fix means the risk persists wherever the device remains reachable.
What to do — in priority order.
- Isolate projector networks — Move Optoma projectors onto a dedicated AV VLAN with ACLs that only permit approved control stations; block user, guest, and server subnets from reaching TCP/2345. For a MEDIUM finding there is no noisgate mitigation SLA, so do this as operational hardening rather than emergency break-fix.
- Block TCP/2345 east-west — Write explicit firewall rules to deny access to port
2345/tcpexcept from named management hosts if business use truly requires it. This is the fastest way to kill the unauthenticated control path when no vendor patch is available. - Remove unnecessary network access — If the projector only receives HDMI input, disconnect Wi-Fi/Ethernet entirely. That closes the exposure completely and is often the cleanest answer for conference-room and signage deployments.
- Inventory and tag exposed Optoma units — Find devices with Optoma banners, MAC OUIs, or open
2345/tcp, then tag them as exception-managed IoT/AV assets. Use that inventory to drive replacement or segmentation within the 365-day remediation window for this MEDIUM verdict. - Monitor for projector API traffic — Alert on internal HTTP requests to
:2345and especiallyPUT /set/patterns. This will not prevent abuse, but it gives you detection for insider misuse and post-compromise east-west activity.
- A standard internet-facing WAF does nothing here; this is usually east-west LAN traffic to an embedded device.
- Endpoint AV/EDR on user laptops does not protect the projector itself; the vulnerable surface is on the embedded Optoma service.
- Waiting for on-device auto-update is not a control; the reviewed disclosure says the March 2026 firmware delivery was incomplete and did not fix this CVE anyway.
Crowdsourced verification payload.
Run this from an auditor workstation on the same network path as the projector, not on the projector itself. Invoke it as bash check_optoma_2345.sh 192.168.1.30; it needs no credentials and no root, only outbound TCP access to the target on port 2345.
#!/usr/bin/env bash
# check_optoma_2345.sh
# Verifies whether an Optoma projector exposes the unauthenticated TCP/2345 control API.
# Usage: bash check_optoma_2345.sh <ip-or-hostname>
# Exit codes:
# 0 = VULNERABLE
# 1 = PATCHED
# 2 = UNKNOWN / error
set -u
TARGET="${1:-}"
PORT="2345"
TIMEOUT="5"
if [[ -z "$TARGET" ]]; then
echo "Usage: bash $0 <ip-or-hostname>"
echo "UNKNOWN"
exit 2
fi
need_cmd() {
command -v "$1" >/dev/null 2>&1
}
if ! need_cmd curl; then
echo "curl is required"
echo "UNKNOWN"
exit 2
fi
# Basic reachability test
if need_cmd nc; then
nc -z -w "$TIMEOUT" "$TARGET" "$PORT" >/dev/null 2>&1
NC_RC=$?
if [[ $NC_RC -ne 0 ]]; then
echo "PATCHED"
exit 1
fi
fi
# Non-destructive read tests first
URLS=(
"http://${TARGET}:${PORT}/get/Volume"
"http://${TARGET}:${PORT}/get/ProjectorID"
"http://${TARGET}:${PORT}/get/Mute"
)
for url in "${URLS[@]}"; do
body_file=$(mktemp)
code=$(curl -m "$TIMEOUT" -sS -o "$body_file" -w "%{http_code}" "$url" 2>/dev/null)
rc=$?
if [[ $rc -eq 0 && "$code" == "200" ]]; then
body=$(tr -d '\r' < "$body_file" | head -c 128)
rm -f "$body_file"
# The published API returns simple scalar values like 0, 1, 93, etc.
if [[ "$body" =~ ^-?[0-9]+$ ]]; then
echo "VULNERABLE"
exit 0
fi
fi
rm -f "$body_file"
done
# If port is open but read probes fail, status is unclear.
# It may be filtered, a different service, or behavior may differ by model/firmware.
if need_cmd nc; then
echo "UNKNOWN"
exit 2
fi
# Fallback if nc was unavailable: infer from curl connectivity failures.
body_file=$(mktemp)
code=$(curl -m "$TIMEOUT" -sS -o "$body_file" -w "%{http_code}" "http://${TARGET}:${PORT}/" 2>/dev/null)
rc=$?
rm -f "$body_file"
if [[ $rc -ne 0 ]]; then
echo "PATCHED"
exit 1
fi
echo "UNKNOWN"
exit 2
If you remember one thing.
2345/tcp, and move them onto an isolated AV segment. For a MEDIUM verdict there is no noisgate mitigation SLA — go straight to the 365-day remediation window; because I found no vendor patch for CVE-2026-30496, your noisgate remediation SLA effectively becomes replacement, retirement, or permanent network isolation of affected units within 365 days, with earlier action for any projector sitting on flat or guest-accessible networks.Sources
What defenders are saying.
Crowdsourced verification outputs.
Results submitted by users who ran the verification payload against their environment.