Getting salty with LLMs: SophosAI unveils new defense against jailbreaking at CAMLIS 2025
On October 22-24, SophosAI will present research on ‘LLM salting’ (a novel countermeasure against jailbreaks) and command line classification at CAMLIS 2025
Similar Posts
Researchers Uncover GPT-5 Jailbreak and Zero-Click AI Agent Attacks Exposing Cloud and IoT Systems
Bys sCybersecurity researchers have uncovered a jailbreak technique to bypass ethical guardrails erected by OpenAI in its latest large language model (LLM) GPT-5 and produce illicit instructions. Generative artificial intelligence (AI) security platform NeuralTrust said it combined a known technique called Echo Chamber with narrative-driven steering to trick the model into producing undesirable
Google Patches Quick Share Vulnerability Enabling Silent File Transfers Without Consent
Bys sCybersecurity researchers have disclosed details of a new vulnerability impacting Google’s Quick Share data transfer utility for Windows that could be exploited to achieve a denial-of-service (DoS) or send arbitrary files to a target’s device without their approval. The flaw, tracked as CVE-2024-10668 (CVSS score: 5.9), is a bypass for two of the 10 shortcomings…
Court clears former Vastaamo CEO Ville Tapio of data protection charges
Bys sThe Helsinki Times reports: The Helsinki Court of Appeal has overturned the criminal conviction of Ville Tapio, the former CEO of psychotherapy provider Vastaamo, in a case linked to one of Finland’s most serious data breaches. The court ruled on Thursday that Tapio was not criminally liable for alleged data protection failures related to the unauthorised access……
Malicious Browser Extensions Infect 722 Users Across Latin America Since Early 2025
Bys sCybersecurity researchers have shed light on a new campaign targeting Brazilian users since the start of 2025 to infect users with a malicious extension for Chromium-based web browsers and siphon user authentication data. “Some of the phishing emails were sent from the servers of compromised companies, increasing the chances of a successful attack,” Positive Technologies…
Chinese Hackers Exploit Ivanti EPMM Bugs in Global Enterprise Network Attacks
Bys sA recently patched pair of security flaws affecting Ivanti Endpoint Manager Mobile (EPMM) software has been exploited by a China-nexus threat actor to target a wide range of sectors across Europe, North America, and the Asia-Pacific region. The vulnerabilities, tracked as CVE-2025-4427 (CVSS score: 5.3) and CVE-2025-4428 (CVSS score: 7.2), could be chained to execute…
WhatsApp, Slack Notifications Could Hijack Google Gemini on Android
Bys sA single poisoned notification from WhatsApp, Slack, SMS, Signal, Instagram, or Messenger could have hijacked Google Gemini’s voice assistant on Android and made it open a victim’s connected windows, fake a message from their boss, push the phone into a Zoom call, or quietly poison its long-term memory. No malicious app on the phone is…