Kaspersky: Limited Effectiveness of ChatGPT to Detect Phishing

Date:

According to research done by Kaspersky, ChatGPT’s effectiveness in detecting malicious phishing links was limited. The accuracy was tested using more than 2000 phishing links that Kaspersky’s anti-phishing software had identified with mixed results. When prompted with “Does this link lead to a phishing website?” and “Is this link safe to visit?”, ChatGPT provided a detection rate of 87.2% and false positive rate of up 23.2%. For the second question, it had a higher detection rate of 93.8% but a false positive rate of 64.3%. These results show that while ChatGPT had the potential to detect phishing links, its accuracy is still too low to be applied in any real-world scenario.

In addition to detection, the experiment also looked into the extent of understanding that ChatGPT had on phishing and its ability to recognize the target. The model recognized the target of a phishing attack in more than 50% of the links it tested. This included brands such as Facebook, TikTok, Google, Amazon and Steam, and multiple banks from around the globe. However, it was found that the security explanations provided by it were often wrong and misleading.

Vladislav Tushkanov, Lead Data Scientist at Kaspersky, commented on ChatGPT as a possible assistant for analysts in fighting phishing attacks. Tushkanov said that language models have their limitations and may not be ready to revolutionize the cybersecurity landscape yet. However, they could still serve as helpful tools for experts.

Kaspersky, founded in 1997 and headquartered in Moscow, is a leading global cybersecurity company that protects users and organizations worldwide with endpoint security, enterprise security, public Wi-Fi security and encryption products. Its anti-phishing, anti-malware, anti-spam and other malware protection products help users maintain online security and protect their data.

See also  Writesonic launches unlimited plan: 20% cheaper than ChatGPT at just $16 a month

Vladislav Tushkanov, Kaspersky’s Lead Data Scientist, is a seasoned machine learning expert with extensive experience in research, development, and productizing of ML algorithms. He focuses on advancing both Kaspersky and the industry by using cutting-edge technologies to develop new artificial intelligence applications for security.

Frequently Asked Questions (FAQs) Related to the Above News

Please note that the FAQs provided on this page are based on the news article published. While we strive to provide accurate and up-to-date information, it is always recommended to consult relevant authorities or professionals before making any decisions or taking action based on the FAQs or the news article.

Share post:

Subscribe

Popular

More like this
Related

Coursera and Microsoft Launch New Entry-Level Professional Certificates for Job Seekers

Coursera and Microsoft launch new entry-level Professional Certificates for job seekers, preparing learners for high-demand jobs in IT, cloud, and business.

Disneyland Cast Members Vote to Unionize with Actors’ Equity Association

Disneyland cast members vote to unionize with Actors' Equity Association, boosting workers' rights in the entertainment industry.

Google Invests in Renewable Energy in Finland, Netherlands, and Belgium

Google invests in renewable energy in Finland, Netherlands, and Belgium to drive AI growth and sustainability, securing long-term wind power contracts.

Reddit Partners with OpenAI for Advanced AI Integration

Enhance your Reddit experience with advanced AI integration from OpenAI. Revolutionizing user interaction on the platform.