CVE-2019-19194 is the SweynTooth "Zero LTK Installation" flaw: the Bluetooth LE pairing code in Telink Semiconductor's BLE SDKs accepts an encryption request in the middle of Secure Connections pairing and encrypts the link with an all-zero key. NVD scores it 8.8, and onsemi says its RSL10 radio was also affected.
Vulnerability summary
CVSS base score
8.8
High severity
EPSS score (FIRST)
1.0%
Probability of exploitation in the wild in the next 30 days
CISA KEV catalog
No
Not listed
Published
Vulnerability ID: CVE-2019-19194 · Source: NVD
Affected components
| Component | CPE | Version range |
|---|---|---|
| Telink Semiconductor BLE SDK (SMP) for TLSR8251, TLSR8253 and TLSR8258 | cpe:2.3:a:telink-semi:tlsr8258_ble_sdk:*:*:*:*:*:*:*:* | through 3.4.0 (NVD); the SweynTooth patch table lists v3.4.0 with the SMP fix |
| Telink Semiconductor BLE SDK (SMP) for TLSR8232 | cpe:2.3:a:telink-semi:tlsr8232_ble_sdk:*:*:*:*:*:*:*:* | through 1.3.0 (NVD); SweynTooth patch table: v1.3.0 (SMP fix) |
| Telink Semiconductor BLE SDK (SMP) for TLSR826x | cpe:2.3:a:telink-semi:tlsr8269_ble_sdk:*:*:*:*:*:*:*:* | through 3.3 (NVD); SweynTooth patch table: v3.3 (SMP fix) |
| onsemi RSL10 Bluetooth radio SDK | — | affected per onsemi customer letter of 2020-05-05; fixed in RSL10 SDK 3.3, released 2020-04-23 |
What does this mean for affected devices?
CVE-2019-19194 is one of the SweynTooth flaws that the ASSET Research Group at SUTD published on 2020-02-11. NVD's description says the Security Manager Protocol (SMP) implementation in Telink Semiconductor BLE SDK versions before November 2019, for TLSR8x5x through 3.4.0, TLSR823x through 1.3.0 and TLSR826x through 3.3, "installs a zero long term key (LTK) if an out-of-order link-layer encryption request is received during Secure Connections pairing". An attacker in radio range "can have arbitrary read/write access to protected GATT service data, cause a device crash, or possibly control a device's function" (NVD). NVD scores it 8.8, vector CVSS:3.1/AV:A/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:H, and records no specific weakness (NVD-CWE-noinfo).
The SweynTooth disclosure explains the attack. A rogue central sends a pairing request that asks for Secure Connections, skips the key exchange and starts link-layer encryption straight away. The Telink peripheral's LTK is still zero, so the attacker can compute the session key, complete encryption and skip key distribution. The researchers say it "affects all products using Telink SMP implementation with support for secure connection enabled" and gives "full communication control over a protected BLE application". As a side note, they point out that affected products that allow only legacy pairing are, for this bug alone, harder to attack. Their tables list the bug for Telink and ON Semiconductor, the affected TLSR8258 SDK 3.4.0 and RSL10 SDK 3.2, and the fixes: Telink TLSR8258 v3.4.0, TLSR8232 v1.3.0 and TLSR826x v3.3 SDKs with an SMP fix, and RSL10 SDK 3.3. They also note that about 480 product listings on the Bluetooth Listing Search used the SoCs they found affected, without claiming each product is exploitable.
onsemi's customer letter of 2020-05-05 confirms that the RSL10 Bluetooth radio is affected by three SweynTooth flaws, Zero LTK Installation among them, and that RSL10 SDK 3.3 with the fixes was released on 2020-04-23. For Zero LTK the letter lists the behaviour on an unpatched SDK as "dependent on customer application", which puts the check on each product maker.
The fix sits in the SoC vendor's SDK, so it reaches a device only when the product maker rebuilds and ships new firmware, the pattern our Bluetooth LE security guide describes for SweynTooth. The CVE is not in the CISA KEV catalog as of 2026-11-16 (catalog version 2026.11.13). FIRST EPSS gives it 0.0101, the 63.7th percentile, in the 2026-11-15 dataset (CISA KEV, FIRST EPSS).
Device makers should find which SoC SDK and version their BLE firmware was built from, move Telink and RSL10 products to the fixed SDKs, and test that the device refuses link-layer encryption before pairing has finished. A firmware SBOM that records the SoC SDK as a component makes this lookup possible later. Other pairing flaws on this site are CVE-2020-26558, a passkey flaw in the Bluetooth specification, and the Zephyr L2CAP bugs such as CVE-2025-10456.
Evidence
- Telink SMP installs a zero LTK on an out-of-order encryption request during Secure Connections pairing; SDK ranges; NVD CVSS 3.1 8.8 AV:A/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:H; NVD-CWE-noinfo. — NVD
- SweynTooth batch 1 published 2020-02-11; Zero LTK attack flow; affects all Telink SMP products with Secure Connections enabled; Telink and ON Semiconductor listed; patch table Telink v3.4.0/v1.3.0/v3.3 SMP fix, RSL10 SDK 3.3; about 480 product listings. — SweynTooth disclosure (ASSET Research Group)
- RSL10 affected by Zero LTK Installation, Channel Map Deadlock and DHCheck Skip; RSL10 SDK 3.3 released 2020-04-23; letter dated 2020-05-05. — onsemi customer letter
- Not in the CISA KEV catalog as of 2026-11-16 (catalog version 2026.11.13). — CISA KEV
- EPSS 0.0101 (63.7th percentile), dataset 2026-11-15, read 2026-11-16. — FIRST EPSS

Related advisories
Sources
Frequently asked questions
Which chips are affected by CVE-2019-19194?
NVD lists Telink Semiconductor BLE SDKs for TLSR8x5x through 3.4.0, TLSR823x through 1.3.0 and TLSR826x through 3.3. The SweynTooth researchers also list ON Semiconductor's RSL10, and onsemi's own letter confirms the RSL10 was affected (per NVD, the SweynTooth disclosure and onsemi).
Does CVE-2019-19194 break BLE encryption itself?
No. The cipher works as designed; the Telink pairing code lets encryption start before Secure Connections pairing has finished and uses an LTK that is still all zeros, so the attacker knows the key (per the SweynTooth disclosure).
How is CVE-2019-19194 fixed in a product?
The SoC vendors fixed their SDKs: Telink SDKs with an SMP fix and onsemi RSL10 SDK 3.3, released 2020-04-23. Each product maker then has to rebuild its firmware on the fixed SDK and ship an update (per the SweynTooth disclosure and onsemi).
Is CVE-2019-19194 exploited?
It is not in the CISA KEV catalog as of 2026-11-16 (catalog version 2026.11.13). FIRST EPSS gives it 0.0101, the 63.7th percentile, in the 2026-11-15 dataset (per CISA KEV and FIRST EPSS).
Questions about this page?