⌛
Loading…
Curated news, CVE analysis, and threat reports from the world's top cybersecurity sources.
HKCERT reports multiple vulnerabilities in internet-facing systems specifically targeting Hong Kong's education sector. The advisory urges affected organizations to patch and secure their systems to prevent exploitation.
HKCERT has received threat intelligence indicating that threat actors are attempting to target Internet-facing systems of Hong Kong’s education sector by scanning for and exploiting their vulnerabilities.
The AI security layer and guardrails for many AI products don't evenly protect against jailbreaking and unsafe actions in every single language.
This article reports that AI security layers and guardrails often fail to protect against jailbreaking and unsafe actions in every language. The uneven coverage creates distinct risks for Europe's multilingual population.
AI safety guardrails, content filters, and moderation systems do not provide equal protection across all languages, creating security gaps that are particularly pronounced in Europe's multilingual environment. Because most AI guardrails are trained disproportionately on English-language data, they are significantly less effective at detecting and blocking harmful content in languages like Maltese, Estonian, Latvian, or even major European languages when prompts mix languages or use code-switching.
HKCERT reported multiple vulnerabilities in MongoDB products. These flaws could allow attackers to compromise systems.