A recent study by MIT Sloan and Safe Security claimed that 80% of ransomwares are now created using generative AI (GenAI). However, several security experts have raised concerns about the study’s findings, calling them “ridiculous” and “made up.” Renowned researcher Kevin Beaumont stated that the study describes many major ransomware groups as using AI without evidence. Other experts, including former hacker Marcus Hutchins, questioned the veracity of the methodology used by the MIT researchers and the definition of AI-powered.
The study was widely cited in publications like Financial Times, but it has since been redacted due to concerns about its accuracy. The researchers have replaced the study with a blank page, stating that they are updating it based on recent reviews. This raises questions about the reliability of such studies, which often rely on small sample groups and rushed reporting.
Experts point out that the definition of AI-powered in the study is too broad and does not accurately reflect real-world use cases. The finding that 95% of enterprises have not seen a return on investment (ROI) from their GenAI spending also seems dramatic and may be exaggerated.
As researchers and institutes face pressure to assess the impact of massive investments in GenAI, such studies can create unnecessary panic rather than reflecting the broader ground reality. This highlights the need for more rigorous research methods and accurate reporting to ensure that conclusions are supported by evidence.
Source: https://cxotoday.com/news-analysis/genai-survey-trap-mits-study-on-ai-ransomware-shows-how-fast-research-can-be-misleading