CVE-2026-20748 is a session-handling flaw in Everon's OCPP back end, api.everon.io: it tied each WebSocket session to the charging station identifier but let several endpoints connect with the same identifier, so the newest connection displaced the real charger and received its commands. CISA covered it in ICSA-26-062-08, and NVD's own score is 8.6.
Vulnerability summary
CVSS base score
8.6
High severity
EPSS score (FIRST)
0.3%
Probability of exploitation in the wild in the next 30 days
CISA KEV catalog
No
Not listed
Published
Vulnerability ID: CVE-2026-20748 · Source: NVD
Affected components
| Component | CPE | Version range |
|---|---|---|
| Everon OCPP back end api.everon.io (OCPP WebSocket session handling for charging stations) | cpe:2.3:a:everon:api.everon.io:*:*:*:*:*:*:*:* | all versions; Everon shut the platform down on 2025-12-01 (CISA ICS Advisory ICSA-26-062-08) |
What does this mean for affected devices?
What does CVE-2026-20748 describe?
According to CISA's ICS Advisory ICSA-26-062-08, "The WebSocket backend uses charging station identifiers to uniquely associate sessions but allows multiple endpoints to connect using the same session identifier. This implementation results in predictable session identifiers and enables session hijacking or shadowing, where the most recent connection displaces the legitimate charging station and receives backend commands intended for that station." The weakness is CWE-613 (insufficient session expiration), and the affected product is api.everon.io, all versions. CISA adds that it may let "unauthorized users to authenticate as other users" or overwhelm the back end with valid session requests.
How is it scored?
CISA scores it 7.3 on CVSS 3.1 (AV:N/AC:L/PR:N/UI:N/S:U/C:L/I:L/A:L) and 6.9 on CVSS 4.0. NVD's own Primary CVSS 3.1 score is 8.6, vector CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:L/I:L/A:H, rating the availability impact high because the real charger is pushed off its session (NVD).
Why is this an OCPP problem and not only an Everon one?
SaiFlow pointed out in 2023 that "the OCPP standard doesn't define how a CSMS should accept new connections from a charge point when there is already an active connection" (SaiFlow). Our OCPP protocol security guide explains why duplicate-connection handling is CSMS behaviour that no OCPP version fixes by itself, and recommends holding both connections and probing with a WebSocket ping or OCPP heartbeat. CISA lists the same CWE-613 finding at other 2026 OCPP back ends, including EVMAPA (CVE-2025-54816 advisory).
What else was in the Everon advisory?
The CSAF record lists it next to CVE-2026-26288 (no authentication on the OCPP WebSocket, CVSS 9.4), CVE-2026-24696 (no limit on authentication requests) and CVE-2026-27027 (station identifiers visible on public maps). The stated mitigation for all four is that "Everon has shut down their platform on December 1st, 2025."
Is it exploited?
It is not in the CISA KEV catalog as of 2026-11-16 (catalog version 2026.11.13). FIRST EPSS gives 0.00292, the 22.1st percentile, in the 2026-11-16 dataset (CISA KEV, FIRST EPSS).
What should CSMS operators do?
Reject or alert on a second connection for an identity that is already connected, authenticate stations with a per-station secret or client certificate, and rate-limit connection attempts. Chargers that used api.everon.io need a new back end.
Evidence
- Sessions keyed on charging station identifiers; multiple endpoints may connect with the same identifier; newest connection displaces the legitimate station (CWE-613); api.everon.io all versions; CISA CVSS 3.1 7.3. — CISA ICS Advisory ICSA-26-062-08
- NVD Primary CVSS 3.1 8.6 (AV:N/AC:L/PR:N/UI:N/S:U/C:L/I:L/A:H). — NVD
- Same advisory lists CVE-2026-26288, CVE-2026-24696 and CVE-2026-27027; mitigation: Everon shut down its platform on 2025-12-01. — CISA CSAF ICSA-26-062-08
- The OCPP standard does not define how a CSMS should accept a new connection when one is already active. — SaiFlow
- Not in the CISA KEV catalog as of 2026-11-16 (catalog version 2026.11.13). — CISA KEV
- EPSS 0.00292 (22.1st percentile) in the 2026-11-16 dataset. — FIRST EPSS

Related advisories
Sources
Frequently asked questions
What happens to the real charger when CVE-2026-20748 is exploited?
CISA says the most recent connection displaces the legitimate charging station and receives the back-end commands intended for it, so the real charger loses its session while an attacker's client takes its place.
Does upgrading to OCPP 2.0.1 prevent this kind of session hijacking?
Not by itself. How a CSMS treats a second connection under an identity that is already connected is implementation behaviour; SaiFlow notes the OCPP standard does not define it. The back end has to reject or verify duplicate connections.
Why is NVD's score higher than CISA's?
CISA rates confidentiality, integrity and availability impact all low (7.3). NVD's Primary assessment rates availability impact high, giving 8.6.
Is CVE-2026-20748 actively exploited?
It is not listed in the CISA KEV catalog as of 2026-11-16 (catalog version 2026.11.13). FIRST EPSS scores it 0.00292, the 22.1st percentile, as of 2026-11-16.
Questions about this page?