Perception Point launches AI model to combat generative AI-based BEC attacks

Perception Point, an web safety platform, revealed its newest innovation to counter the rising tide of AI-generated electronic mail threats. The firm’s new detection expertise employs AI-powered massive language fashions (LLMs) and deep studying structure to determine and thwart enterprise electronic mail compromise (BEC) attacks facilitated by generative AI applied sciences.

Criminals are exploiting generative AI expertise to perform subtle, exactly focused attacks towards organizations of all sizes. The expertise has emerged as a brand new potent device for cybercrime, particularly in social engineering and BEC attacks, because it permits the creation of high-quality, personalised emails that resemble human output.

According to Verizon’s current knowledge breach investigation report, over 50% of social engineering incidents might be attributed to BEC. Perception Point’s 2023 annual report additionally reveals an 83% surge in BEC makes an attempt.

To tackle this escalating menace, the corporate has developed an modern detection model primarily based on LLMs, which make the most of transformers — AI fashions able to comprehending the semantic context of the textual content, related to famend LLMs corresponding to OpenAI’s ChatGPT and Google’s Bard. 

The answer can subsequently determine distinct patterns in LLM-generated textual content, a vital consider detecting and thwarting gen AI-based threats.

Beyond legacy safety options

Perception Point asserts that typical safety distributors typically fail to obtain the required stage of detection accuracy by means of contextual and behavioral evaluation.

The firm states that whereas superior electronic mail safety methods use contextual and behavioral detection, they nonetheless battle to determine the newly enhanced attacks facilitated by generative AI. This is as a result of these attacks circumvent the standard patterns that the detection strategies had been initially designed to acknowledge.

Moreover, the corporate claims that options at the moment accessible available in the market rely solely on post-delivery detection. That means the malicious electronic mail can sit within the person’s inbox for an prolonged interval earlier than being eliminated.

“Legacy email security solutions which rely on signatures and reputation analysis struggle to stop even the most basic payload-less BEC attacks,” Tal Zamir, CTO of Perception Point, advised VentureBeat. “Our new model’s key strength lies in recognizing the repetition of identifiable patterns in LLM-generated text. The model uses a unique three-phase architecture that detects BEC at the highest detection rates and minimizes false positives.”

Zamir mentioned the answer’s distinction lies in its complete scanning of all emails, quarantining these recognized as malicious earlier than they attain the person’s inbox. He defined that this proactive strategy eliminates the dangers and potential damages related to detection-based strategies that depend on figuring out and addressing threats as soon as they’ve infiltrated the system.

Additionally, the answer incorporates a managed incident response service, relieving clients’ SOC groups of the duty to swiftly reply to incidents and deploy new algorithms in actual time to counter novel and rising threats.

Perception Point claims its model reveals distinctive pace in processing incoming emails, with a median time of 0.06 seconds. The model was initially educated on a whole lot of 1000’s of malicious samples captured by the corporate and is constantly up to date with new knowledge to optimize its effectiveness.

Leveraging generative AI to decrease email-based attacks

Perception Point’s Zamir mentioned the brand new attacks embody cybercriminals exploiting pretend emails to impersonate trusted organizations. Using social engineering methods, the attackers deceive staff into transferring massive sums of cash or disclosing confidential knowledge.

“Attackers exploit the fact that employees in the modern enterprise are the weakest link in the organization regarding security,” Zamir advised VentureBeat. “They are leveraging BEC text-based attacks, which normally do not have malicious payloads such as URLs or malicious files, and thus bypass traditional email security systems, arriving into the users’ inboxes.”

He additional said that the emergence of generative AI, particularly LLMs, has given a lift to impersonation, phishing and BEC attacks. This development empowers cybercriminals to function at larger pace and scale than ever earlier than.

“Tasks that once required extensive time and effort, such as target research, reconnaissance, copywriting and design, can now be accomplished within minutes using carefully crafted prompts,” mentioned Zamir. “This amplifies the threat by expanding the pool of potential victims and significantly increasing the chances of successful attacks.”

To cut back false positives that come up from the in depth use of generative AI for respectable emails, Perception Point makes use of a particular three-phase structure in its model. 

Following an preliminary scoring course of, the model employs transformers and clustering algorithms to categorize electronic mail content material. By integrating insights from these phases with supplementary knowledge, corresponding to sender fame and authentication protocol info, the model predicts whether or not an electronic mail is AI-generated and determines if it presents a possible menace.

“Our model dynamically scans every email, including the embedded URLs and files, with a patented HAP (Hardware Assisted Platform) detection layer. This is our proprietary next-gen sandbox that dynamically scans content at the CPU/memory level,” mentioned Zamir.

What’s subsequent for Perception Point?

Zamir mentioned that his firm goals to develop AI capabilities to sift by means of huge quantities of information, figuring out potential threats and offering clients with actionable intelligence.

He emphasised that integration of generative AI bots into collaboration apps like Slack or Teams, browsers like Edge, and cloud storage companies like Google Drive or OneDrive has created new avenues for potential attacks.

“Perception Point recognizes these emerging threats, and we are developing AI security solutions designed to prevent, detect and respond to the ever-increasing threat landscape complexity,” mentioned Zamir. “We will continue to ensure that our clients can leverage the power of generative AI without compromising their security posture.”

VentureBeat’s mission is to be a digital city sq. for technical decision-makers to achieve data about transformative enterprise expertise and transact. Discover our Briefings.