CVE-2026-104074 Coturn 4.10.0 Uninitialized Stack Memory Disclosure via ERROR-CODE

Coturn 4.10.0 contains an uninitialized memory disclosure vulnerability that allows remote unauthenticated attackers to leak stack memory contents by sending a TURN Allocate request without credentials. Attackers can exploit the stuniniterrorresponsecommonstr function in src/client/nsturnmsg.c,...

EUVD-2026-94363

Coturn 4.10.0 contains an uninitialized memory disclosure vulnerability that allows remote unauthenticated attackers to leak stack memory contents by sending a TURN Allocate request without credentials. Attackers can exploit the stuniniterrorresponsecommonstr function in src/client/nsturnmsg.c,...

CVE-2026-104074: Use of Uninitialized Resource

Coturn 4.10.0 contains an uninitialized memory disclosure vulnerability that allows remote unauthenticated attackers to leak stack memory contents by sending a TURN Allocate request without credentials. Attackers can exploit the stuniniterrorresponsecommonstr function in src/client/nsturnmsg.c,...

PT-2026-107529

Coturn 4.10.0 contains an uninitialized memory disclosure vulnerability that allows remote unauthenticated attackers to leak stack memory contents by sending a TURN Allocate request without credentials. Attackers can exploit the stun init error response common str function in src/client/ns turn...

vLLM: Harmony tool continuations drop `cache_salt` — restoring a cross-tenant prefix-cache membership oracle

Affected - Ecosystem / package: pip / vllm - Affected versions: vLLM ≤ 0.25.1 confirmed on 0.25.1, commit 752a3a504485. The lower bound predates 0.25.1; maintainers can confirm how far back the tool-continuation re-submission has omitted the salt. Summary On the GPT-OSS "Harmony" path POST...

The Model Plants the Trigger: Answer-Side Backdoor Attacks in Multi-Turn Large Language Models

Safety alignment in Large Language Models LLMs remains vulnerable to backdoor attacks. Existing LLM backdoors are almost all input-centric: activation depends on explicit trigger patterns in the user input, so modern guardrails are built to sanitize the input space. We challenge this assumption...

CVE-2026-100251

Wormhole.app as deployed before 2026-08-22 misconfigures the coturn TURN server and does not properly restrict TCP relay peers, allowing an unauthenticated attacker to access instance metadata or to source TCP connections from the Wormhole relay's IP...

CVE-2026-100251

Versions of Wormhole.app deployed before 2026-08-22 are vulnerable to a server-side request forgery (SSRF) due to a misconfiguration in the coturn TURN server . The flaw arises because the application fails to properly restrict TCP relay peers , enabling an unauthenticated attacker to access inst...

EUVD-2026-91032

Wormhole.app as deployed before 2026-08-22 misconfigures the coturn TURN server and does not properly restrict TCP relay peers, allowing an unauthenticated attacker to access instance metadata or to source TCP connections from the Wormhole relay's IP...

CVE-2026-100251 Wormhole.app SSRF

Wormhole.app as deployed before 2026-08-22 misconfigures the coturn TURN server and does not properly restrict TCP relay peers, allowing an unauthenticated attacker to access instance metadata or to source TCP connections from the Wormhole relay's IP...

CVE-2026-100251 Wormhole.app SSRF

Wormhole.app as deployed before 2026-08-22 misconfigures the coturn TURN server and does not properly restrict TCP relay peers, allowing an unauthenticated attacker to access instance metadata or to source TCP connections from the Wormhole relay's IP...

PT-2026-103979

Wormhole.app as deployed before 2026-08-22 misconfigures the coturn TURN server and does not properly restrict TCP relay peers, allowing an unauthenticated attacker to access instance metadata or to source TCP connections from the Wormhole relay's IP...

CVE-2026-100251: Server-Side Request Forgery (SSRF)

Wormhole.app as deployed before 2026-08-22 misconfigures the coturn TURN server and does not properly restrict TCP relay peers, allowing an unauthenticated attacker to access instance metadata or to source TCP connections from the Wormhole relay's IP...

CLEANSTART-2023-MJ86716 Coturn is free open source implementation of TURN and STUN Server

Security vulnerability affects the coturn package. Coturn is free open source implementation of TURN and STUN Server...

Speculative Safety Honeypot: Toward Proactive Defense against Multi-Turn Agent Attacks

As Large Language Model LLM agents are increasingly deployed in complex environments, multi-turn interaction attacks have become a significant security challenge. Existing detection methods typically rely on historical context. However, this retrospective logic struggles to identify deep maliciou...

The Geometry of Harmfulness in Multi-Turn Attacks

Large language models LLMs remain vulnerable to adversarial attacks that circumvent safety alignment to elicit harmful outputs. It remains unclear how harmfulness and refusal representations evolve over the course of multi-turn attacks, and why single-turn defenses are less effective in multi-tur...

Does the Unsafe Gradient Survive a Conversation? on the Fragility of Gradient-Based Jailbreak Detection in Multi-Turn Dialogue

Safety-aligned language models are commonly deployed as multi-turn assistants, which lets adversaries spread unsafe intent across several user turns instead of a single prompt. Gradient-based jailbreak detectors such as GradSafe were developed for single prompts: they score an input by the...

[SECURITY] Fedora 45 Update: baresip-4.11.0-1.fc45

A modular SIP user-agent with support for audio and video, and many IETF standards such as SIP, SDP, RTP/RTCP and STUN/TURN/ICE for both, IPv4 and IPv6. Additional modules provide support for audio codecs like Codec2, G.711, G.722, GSM, L16 and Opus, audio drivers like ALSA, GStreamer, JACK Audio...

EvoFlint: An Evolutionary Atlas of Multi-Turn LLM Vulnerabilities

Frontier language models that refuse harmful single-turn prompts often comply when the same intent is reached gradually over many turns, making multi-turn attacks one of the least understood failure modes of large language models. Most automated red-teaming methods treat this as a generation...

SUSE CVE-2026-40613

Coturn is a free open source implementation of TURN and STUN Server. Prior to 4.10.0, the STUN/TURN attribute parsing functions in coturn perform unsafe pointer casts from uint8t to uint16t without alignment checks. When processing a crafted STUN message with odd-aligned attribute boundaries, thi...