Leading cybersecurity provider, Perception Point, has unveiled its latest innovation designed to combat the increasing threat of AI-generated email attacks. The company’s AI-powered technology utilizes Large Language Models (LLMs) and Deep Learning architecture to effectively detect and prevent Business Email Compromise (BEC) attacks, which have become more sophisticated due to the rise of Generative AI (GenAI) technologies.
In recent years, threat actors have been exploiting evolving GenAI technology to execute highly targeted and advanced attacks against organizations of all sizes. Democratized capabilities, including the creation of personalized emails that resemble human-like communication, have turned GenAI into a powerful tool for cybercrime. Social engineering and BEC attacks, specifically, have witnessed a substantial surge in activity in the last year. According to the DBIR 2023 Report, BEC accounted for more than 50% of incidents involving social engineering, while Perception Point’s 2023 Annual Report highlighted an 83% growth in BEC attempts.
Perception Point’s response to this escalating threat is a detection model based on LLMs, harnessing the power of Transformers- AI models capable of understanding the semantic context of text. This approach allows the solution to identify unique patterns in LLM-generated text, a crucial factor in detecting and thwarting GenAI-based threats. Traditional security vendors often struggle to achieve this using contextual and behavioral analysis.
The LLM-based model developed by Perception Point processes incoming emails at an impressive average speed of 0.06 seconds, aligning with the company’s ability to dynamically scan 100% of content in near real-time. Initially trained on hundreds of thousands of malicious samples captured by the company, the model continues to be updated with new data to enhance its effectiveness.
Tal Zamir, CTO of Perception Point, emphasized the urgent need for cutting-edge defenses against GenAI-powered threats in an increasingly complex threat landscape. Zamir stated, “By reversing this dynamic and proactively leveraging AI for detection, we are able to prevent these threats before they even reach the user’s inbox – a paradigm shift in the fight against BEC attacks.”
To minimize false positives resulting from the use of generative AI for legitimate emails, Perception Point implemented a unique 3-phase architecture. After initial scoring, the model categorizes the email content using Transformers and clustering algorithms. It then integrates insights from these steps with additional data like sender reputation and authentication protocol information. This integrated process enables the model to accurately predict whether an email is AI-generated and if it potentially poses a threat.
Perception Point is leading the charge against GenAI-generated content attacks with its AI-powered cybersecurity solution. Leveraging patterns in LLM-generated content, advanced image recognition, anti-evasion algorithms, and patented dynamic engines, the company offers a robust and proactive defense against threats, neutralizing them before they reach users.
For more information on Perception Point’s pioneering approach to combating GenAI-generated content attacks, visit their website.
About Perception Point:
Perception Point is a Prevention-as-a-Service company that provides fast and accurate next-generation detection and response to all attacks across email, cloud collaboration channels, and web browsers. The company’s natively integrated incident response service acts as a force multiplier to SOC teams, reducing management overhead, improving user experience, and delivering continuous insights. Their cloud-native service, which can be deployed in minutes without any changes to the enterprise’s infrastructure, replaces legacy systems to prevent various advanced attacks such as phishing, BEC, spam, malware, Zero-days, and ATO attacks. Fortune 500 enterprises and organizations worldwide trust Perception Point’s solution to defend against content-borne attacks across email and cloud collaboration channels.