Zscaler's 2025 AI Security Report found that enterprise use of artificial intelligence and machine learning has grown more than 30-fold over the past year, leading to increased cybersecurity risks, SiliconAngle reports.
The study, based on 536 billion AI transactions within the Zscaler Zero Trust Exchange, reveals that companies are sending 4,500 terabytes of data to AI platforms while blocking nearly 60% of AI-related transactions to mitigate threats such as data leakage, unauthorized access, and compliance violations. ChatGPT leads AI adoption with 47.1% of global enterprise AI usage, yet it is also the most frequently blocked due to concerns over sensitive data exposure. Other commonly restricted applications include Grammarly, Microsoft Copilot, QuillBot, and Wordtune. The report also highlights emerging security risks linked to agentic AI and Chinas DeepSeek, which could enable more sophisticated cyberattacks. The finance and insurance sector leads in AI usage with 28.4% of enterprise traffic, followed by manufacturing at 21.6% and services at 18.5%. Additionally, the report named the US and India as the top adopters of AI.
