A quarter of European organisations have blocked Grok AI, a chatbot developed by Elon Musk’s xAI, according to new research by Netskope Threat Labs. Integrated into its sister platform X (formerly Twitter) and promoted for its “sense of humour,” Grok AI has become a topic of controversy due to contentious responses. Recently, the Irish Data Protection Commission (DPC) initiated an investigation into how personal data from publicly accessible posts on X, made by EU users, is processed for training generative AI (GenAI) models, specifically targeting the Grok large language models (LLMs).
The Netskope report further identifies other frequently blocked GenAI applications within the region. Stable Diffusion is restricted by 41% of organisations, AiChatting by 31%, and NotebookLM by 26%. Privacy and licensing issues predominantly drive these restrictions, leading many organisations to favour more secure or compliant alternatives. For instance, while ChatGPT is blocked by just 9.8% of companies, this indicates a strategic selection of tools perceived as lower risk.
GenAI applications facing restrictions across Europe
Europe demonstrates significant adoption of cloud-based GenAI tools, with 91% of organisations integrating them into operations. Notably, 97% of these entities utilise tools incorporating GenAI-powered features, and 96% rely on applications that leverage user data for training purposes. These figures align with global trends where 94% of organisations have adopted GenAI apps in the cloud.
Interestingly, there has been a marked reduction in the use of personal GenAI accounts, declining from 73% to 58% over the past year. This shift underscores a move towards company-approved solutions offering enhanced control and safeguarding of sensitive data.
Globally and within Europe, ChatGPT remains the leading GenAI app, used by 79% of organisations in the region. Google Gemini is also gaining traction as a notable alternative. Other widely used GenAI tools include GitHub Copilot at 41%, Perplexity AI at 35%, and Microsoft 365 Copilot at 34%. These tools are becoming integral to enhancing productivity across various domains.
The extensive deployment of GenAI applications necessitates robust data loss prevention (DLP) tools to manage emerging security threats. Sensitive data, such as source code and information regulated under GDPR, are frequently shared with these platforms, posing substantial compliance challenges. Organisations face potential financial penalties if they do not handle this data responsibly.
Despite inherent security risks, employees continue to engage with personal GenAI apps, often risking exposure of critical information like source code. This trend suggests developers use these tools to boost productivity even if it means bypassing officially sanctioned platforms. Consequently, it is imperative for organisations to establish clear guidelines and oversight on using GenAI tools beyond corporate control.
Read more: 57% of enterprise employees input confidential data into AI tools, survey reveals
