{"id":19012,"date":"2023-12-07T08:41:00","date_gmt":"2023-12-07T07:41:00","guid":{"rendered":"https:\/\/www.m-q.ch\/?p=19012"},"modified":"2023-12-07T08:41:00","modified_gmt":"2023-12-07T07:41:00","slug":"ki-supported-cyber-attacks-on-the-rise","status":"publish","type":"post","link":"https:\/\/www.m-q.ch\/en\/ki-supported-cyber-attacks-on-the-rise\/","title":{"rendered":"AI-assisted cyberattacks on the rise"},"content":{"rendered":"
\"\"
Trend Micro warns of the transformative role of generative artificial intelligence. (Image: www.unsplash.com)<\/figcaption><\/figure>\n

The widespread availability and improved quality of generative AI combined with Generative Adversarial Networks (GANs) to generate realistic photo, audio and video content will have a lasting impact on the phishing landscape in 2024. Trend Micro predicts a new wave of Business Email Compromise (BEC), Virtual Kidnapping and other scams - triggered by the cost-effective creation of such content.<\/p>\n

With lucrative profit prospects, threat actors for such campaigns either use legal AI tools with stolen credentials and VPNs to hide their identities or develop their own malicious generative AI tools. But AI models themselves will also come under attack in 2024: While the data sets of generative AI and LLMs (Large Language Models) are difficult for threat actors to influence, specialized cloud-based machine learning models represent an attractive target. They are trained with more specific data sets and can fall victim to data poisoning attacks - from exfiltrating sensitive data to disrupting fraud filters and even interfering with connected vehicles. Such attacks already cost actors less than 100 US dollars today.<\/p>\n

\"Advanced LLMs that speak any language pose a significant threat as they avoid common clues for phishing attacks, such as unusual formatting or grammatical errors. This makes it more difficult to detect such attacks,\" reports Udo Schneider, IoT Security Evangelist Europe at Trend Micro. \"Companies must therefore adapt their existing phishing training and also introduce modern technical protection measures. Advanced defenses not only outperform human detection capabilities, but also ensure resilience against these attack tactics.\"<\/p>\n

Such security developments may in turn lead to increased scrutiny by regulators and also prompt the technology industry to take matters into its own hands: \"In the coming year, the cyber industry will overtake legislators in the development of cybersecurity-specific AI guidelines. The industry is quickly moving towards voluntary self-regulation,\" continues Udo Schneider.<\/p>\n

The Japanese security expert Trend Micro also names other developments that IT security managers should pay particular attention to in 2024:\u00a0<\/p>\n