SoftBank has developed Cloud Proxy, a custom AI security and management gateway that enables its enterprise vision of deploying 100 autonomous AI agents per person. Starting from 2023, the system manages API keys, user identity verification, and token consumption across multiple AI models while preventing unauthorized access. The architecture has demonstrated the ability to handle dramatically increased request volumes within a 14-week scaling window, suggesting the technical foundation needed to support large-scale enterprise agent deployment.
Summaries like this, in your inbox every morning.
Sign up free →What happened
SoftBank has developed Cloud Proxy, an internal AI gateway system designed to support its vision of "100 agents per person." The gateway, which began development in 2023, manages security, identity verification, and token usage across multiple AI models including Azure OpenAI Service, Gemini, and Sarashina.
Why it matters
The system solves a core security and operational challenge in deploying many autonomous AI agents (software that makes decisions and takes actions) simultaneously. By handling API key management, preventing unauthorized access, and enabling fine-grained monitoring of token usage—the basic units measured for AI processing—Cloud Proxy allows SoftBank to scale agent deployment while maintaining security and controlling costs.
What to watch
SoftBank reports that its scaling tests showed the system can support "100 agents per person" architecture in 14 weeks, with request volumes doubling while maintaining infrastructure stability. The company is adding support for new AI models through a weekly update process and is aiming to demonstrate that integration with different identity providers can reduce enterprise AI adoption friction by 70 percentage points.
SoftBank's Cloud Proxy represents a response to a concrete operational challenge that emerges when enterprises move from single-AI-system deployments to managing dozens or hundreds of autonomous agents simultaneously. The article reveals that the company identified five key "pillars" of risk—licensing, security, operation, governance, and compatibility—before designing a unified gateway to mediate between internal tenants and external AI services. By centralizing identity verification, API key management, and usage tracking, the system shifts security responsibility from individual teams to a single hardened layer, addressing what SoftBank describes as a core blocking point for enterprise adoption.
The scaling story is particularly concrete: the system moved from a 14-week optimization window to supporting 5-minute scaleouts in later testing, a pattern SoftBank attributes to a tool called "n8n" for workflow automation. This progression suggests the architecture is moving from prototype to operational readiness. The company also notes that adding support for new models now takes roughly one week rather than requiring extended integration work, a measure of platform maturity. These technical gains matter because they directly reduce the friction barrier cited in the article: if identity provider integration alone can drop adoption obstacles by 70 percentage points, a system that handles identity, security, and multiple models in one place may reduce institutional deployment time significantly.
The vision of "100 agents per person" remains a strategic framing rather than a near-term shipping date, but Cloud Proxy's role is to make it technically feasible by solving cost, security, and operational control at scale. SoftBank's teams are using this infrastructure internally to test the agent architecture while external deployments remain ongoing.
No comments yet. Be the first to share your thoughts!
Log in to join the discussion
Get curated AI news from 200+ sources delivered daily to your inbox. Free to use.
Get Started FreeFree · takes 30 seconds · unsubscribe anytime
1 minute a day. The AI essentials.
200+ sources · Email / LINE / Slack
