AI security shifts toward continuous testing as Cresta establishes a Japan base

Developments

Japan’s enterprise-AI market is becoming more operationally specific. One new service treats security testing as a recurring process tied to remediation, while a US customer-experience vendor is building a formal local base. Both developments are company announcements, so they establish strategy and availability rather than independently measured results.

1. TWOSTONE Sync and Layer8 package penetration testing as a recurring service

TWOSTONE Sync announced a partnership with Layer8 to offer penetration testing for web applications and APIs. The release describes an AI agent exploring possible attack paths, security engineers reproducing and verifying findings, and a monthly cycle combining one test with one remediation check. TWOSTONE Sync also says engineering resources can be supplied when fixes require additional capacity. The operating model is the notable development; claims about coverage and effectiveness still require buyer-side evaluation. (Read the announcement)

2. Cresta formalizes its Japan push with a local company and country manager

Cresta announced a Japanese entity in Marunouchi and named Hajime Tamura as country manager. The company describes its offer as an enterprise customer-experience AI platform combining automated agents, real-time operator assistance and conversation analysis. A local entity can make sales, implementation and Japanese-language support easier to organize, but the release does not independently establish adoption, integration quality or performance in Japan. (Read the announcement)

What the two moves suggest

Enterprise AI is shifting from broad experimentation toward named operating functions and ongoing service relationships. Buyers should still separate announced capabilities from evidence about deployment quality, security controls, integration cost and measurable results.