⌛
Loading…
Curated news, CVE analysis, and threat reports from the world's top cybersecurity sources.
This article discusses new research showing that AI browsers from leading vendors are still vulnerable to prompt injection attacks, even after deploying various security measures. The findings indicate that no perfect fix exists, leaving users exposed to potential manipulation of AI agents through crafted prompts.
A Canadian man pleaded guilty to participating in a scheme to access company accounts at cloud storage provider Snowflake and steal data from at least 165 organizations. The attack was designed to extort millions of dollars from victims.
AWS has introduced AWS Continuum, a tool that uses multiple AI models to secure code by discovering vulnerabilities and performing contextual prioritization, validation, and remediation. New partnerships with Anthropic and OpenAI extend Continuum directly into developer workflows through Claude Code, OpenAI Codex, and Kiro.
AWS Continuum helps secure code at machine speed by combining multiple AI models to discover vulnerabilities and perform contextual prioritization, validation, and remediation. New partnerships with Anthropic and OpenAI extend Continuum directly into developer workflows through Claude Code, OpenAI Codex, and Kiro.
Hackers exploited a SQL injection vulnerability to install a post-exploitation toolkit directly inside an Oracle database. This toolkit was then used to breach a corporate network.
Researchers warn that CSS, once only used for design, is now powerful enough to exfiltrate data from webmail inboxes. Some vendors are not prepared for this hidden threat.