WASHINGTON (dpa-AFX) - Google has revealed that its flagship artificial intelligence chatbot, Google Gemini, has been the target of repeated attempts by 'commercially motivated' actors to clone the system.
These attacks involve bombarding the chatbot with large volumes of carefully crafted prompts in an effort to uncover the patterns and logic underlying its responses.
According to a report published by Google, the company has faced a rise in so-called 'distillation attacks,' also known as model extraction. The goal of these attacks appears to be replicating or enhancing competing AI systems by reverse-engineering Google's proprietary technology.
Google believes the activity is largely driven by private firms or researchers seeking a competitive edge, though the company declined to identify specific suspects or locations.
John Hultquist, chief analyst at Google's Threat Intelligence Group, warned that such attacks are likely to become more common, particularly as smaller companies deploy customized AI tools trained on valuable or sensitive data.
Major AI developers, including OpenAI, the creator of ChatGPT, have invested billions of dollars in building large language models and consider their internal architectures highly confidential.
OpenAI previously accused Chinese startup DeepSeek of using distillation techniques to improve its own models.
Copyright(c) 2026 RTTNews.com. All Rights Reserved
Copyright RTT News/dpa-AFX
© 2026 AFX News
