Varonis Threat Labs has exposed a long-standing vulnerability in web browsers related to how mixed text directions are displayed. Known as the BiDi Swap flaw, it leverages inconsistencies in the way browsers process Left-to-Right (LTR) and Right-to-Left (RTL) scripts. This loophole allows attackers to create URLs that appear legitimate but actually lead to malicious websites. By blending characters from different scripts, attackers can visually mimic trusted domains, making phishing attempts far more convincing. At the core of this issue is the Unicode Bidirectional Algorithm, which was developed to ensure proper rendering of text that mixes languages like English and Arabic. While it functions effectively for most domain names, Varonis discovered that it becomes unreliable when handling subdomains and URL parameters. This opens a path for manipulation, where malicious links can look safe but behave unpredictably. Attackers exploit this by crafting URLs that take advantage of subtle visual misdirection. In addition to the BiDi Swap technique, hackers often rely on related spoofing methods. These include Punycode-based homograph attacks, where visually similar characters from different alphabets are substituted (e.g., replacing “a” with the Cyrillic equivalent), and RTL override tricks that disguise file extensions, turning harmful files into seemingly safe ones. These methods all capitalize on minor quirks in text rendering to bypass user suspicion and browser detection. Despite the problem being recognized for over a decade, a universal fix has yet to be implemented. Chrome offers limited protection by flagging some lookalike domains, while Firefox attempts to highlight the main domain in the address bar. Microsoft Edge claims the issue is resolved, though URL formatting remains unchanged. Varonis urges users to stay cautious when inspecting URLs and recommends that browser developers improve domain clarity and detection features to help prevent exploitation.
Researchers at Cofense Intelligence found the campaign of group Lone None distributing advanced malware in the form of spoofed copyright takedown notices. The campaign delivers the...
A widespread security issue has been uncovered in mobile apps using Firebase, where misconfigured services expose sensitive user data to unauthenticated attackers. First disclosed ...
A new phishing campaign is targeting PyPI maintainers using deceptive emails that mimic official PyPI communications. These emails, crafted to look like account verification reques...