VirusTotal, a leading cybersecurity platform, recently introduced a groundbreaking update to its Crowdsourced AI project by integrating a cutting-edge AI model specifically designed to analyze suspicious macros in Microsoft Office files. Developed by Dr. Ran Dubin from Ariel University and ByteDefend Cyber Lab, this new AI model is a significant addition to the platform’s capabilities when it comes to detecting and assessing potential threats in Word, Excel, and PowerPoint documents.

The Crowdsourced AI initiative by VirusTotal harnesses the power of multiple AI models and community contributions to enhance cyber defense strategies. While AI-based models might not be flawless, they play a crucial role in collaboration with other technologies to identify and evaluate emerging cybersecurity risks effectively.

The incorporation of ByteDefend’s AI model enhances VirusTotal’s existing Code Insight features, which already utilize three separate AI systems dedicated to Microsoft Office files. This consolidation of AI technologies bolsters the platform’s ability to detect and analyze malicious content embedded within various types of documents.

One recent instance showcased the collective agreement of all three AI models on the malicious nature of an XLS file under examination, albeit with varying degrees of detail provided by each model. This unanimous decision underscores the value of employing multiple AI engines to conduct comprehensive threat assessments for improved accuracy and efficiency.

On a separate occasion, ByteDefend flagged a DOC file as malicious, while another system from Hispasec deemed it safe. Such discrepancies in threat analysis highlight the subjective nature of cybersecurity evaluations and underscore the importance of context in decision-making processes. Despite discrepancies in conclusions, AI models offer detailed insights into the functionality of macros, empowering human researchers to make informed decisions based on the information provided.

Enhancements in the search capabilities within VirusTotal’s platform allow users to access AI-generated reports through VT Intelligence. Specific modifiers such as “bytedefend_ai_analysis:” enable users to search for AI analysis outputs, while “bytedefend_ai_verdict:” facilitates searches based on verdicts. For example, users can query the platform for ByteDefend reports mentioning a specific term like “telegram” with a malicious judgment.

The significant contributions of ByteDefend Cyber Lab and Dr. Ran Dubin in advancing cybersecurity technologies have been acknowledged and appreciated by VirusTotal. The platform aims to expand its Crowdsourced AI project by engaging a diverse pool of individuals possessing various skills and expertise to collectively strengthen defense mechanisms against evolving cyber threats.

By integrating ByteDefend’s AI model, VirusTotal continues to enhance its capabilities in identifying and analyzing threats within Microsoft Office files, solidifying its position as a frontrunner in cybersecurity innovation. The platform encourages professionals in the security domain to join forces in this collective effort to bolster cybersecurity defenses and tackle the ever-changing landscape of online threats.

Source link