Tokyo-based Tempura Technologies, a cybersecurity startup, said on November 12 that it has joined Nvidia Inception and will evaluate Nvidia tools for Security, its AI security service.

The company said it will test Nvidia Nim and NeMo Guardrails for closed AI environments that let companies use generative AI without sending internal data to external AI services.

It also plans to use Nvidia’s garak tool for simulated attacks on AI features such as chatbots and assistants.

Inception offers access to Nvidia technology, technical training, consulting, and market support.

Japan’s National Police Agency reported 226 ransomware cases in 2025. The Information-technology Promotion Agency’s 2026 list of organizational security threats included AI-related cyber risks for the first time.

Japanese government guidance has urged companies not to enter trade secrets and other highly confidential information into external AI services without safeguards. A 2023 survey found over 60% of Japanese companies were using or considering generative AI, with information leaks among their concerns.

🔗 Source: Tempura technologies Inc.