Lasso Security integration with Truefoundry AI Gateway

Every enterprise wants to ship AI applications faster. Developers want to plug into the best models and move quickly. Security teams want to be sure that every prompt and every response is safe, compliant, and free of leaked data. These two goals usually pull in opposite directions, and the friction only grows as more teams build more applications on more models. The big question is simple: how do you protect every AI interaction across your organization without bolting custom security code onto each application one by one?

That is exactly the problem this partnership solves. We are excited to announce the integration of Lasso Security with the TrueFoundry AI Gateway. Together they let you enforce consistent, enterprise-grade security across every model and every application from one central place, so your teams can keep building at full speed while your security posture stays intact.

TrueFoundry AI Gateway as a Central Control Plane

The TrueFoundry AI Gateway is the command center for enterprise AI. It gives developers and platform teams unified access to hundreds of large language models, deep observability into every request, and smart routing and fallback all in one place. Instead of wiring each application to a different provider and maintaining a tangle of API keys and SDKs, teams point everything at the Gateway and let it handle the rest.

As AI adoption scales, the real challenge is no longer getting access to models. It is managing the complexity that comes after: multiple providers, shifting APIs, and strict compliance requirements. The Gateway brings order to that complexity by unifying access, enforcing policy, and giving you full visibility across every model and environment. Security that lives at this layer protects every application at once, with no per-app engineering required.

Lasso Security: Purpose-Built Protection for GenAI

Lasso Security is an AI security platform built specifically for the way large language models behave. Rather than retrofitting traditional security tools, Lasso understands the intent behind a prompt and a response, which lets it catch the threats that generic filters miss.

Lasso protects against the full spectrum of LLM risk. It detects and blocks prompt injection and jailbreak attempts before they manipulate your models, with protections mapped to the OWASP Top 10 for LLMs. Intent classifiers infer what a prompt is actually trying to do, catching adversarial inputs that slip past keyword filters. You define custom policies using an LLM-as-a-judge approach, and those policies act as deputies inside the Lasso platform that travel with every request. It also prevents data leakage by masking PII such as emails and phone numbers on the fly, flags toxic content, and gives security teams visibility into shadow AI usage across the organization.  

Better Together: One Integration, Centralized Control

Bringing Lasso Security and the TrueFoundry AI Gateway together makes securing your AI both simpler and stronger. Instead of embedding Lasso into every application with separate SDKs and scattered code changes, you connect it once at the Gateway and apply it to any model behind it.

The Gateway listens to Lasso's verdict on every call and enforces it automatically, whether that means blocking a malicious prompt or masking sensitive data before it ever reaches a model. You configure Lasso as a guardrail directly in the TrueFoundry dashboard, which gives you one place to manage security policy across teams, models, and environments. And because Lasso's findings are paired with the Gateway's detailed request tracking, you get a complete audit trail that makes compliance and incident response far easier.

How the Lasso Security and TrueFoundry Integration Works

The integration introduces a layered security approach around every request.

1. When your application sends a request, it first passes through the TrueFoundry AI Gateway.
2. The Gateway forwards the prompt to Lasso Security, which classifies it against your deputies and policies.
3. If Lasso returns an allow verdict, the request proceeds safely to the model. If it returns a block verdict, the Gateway halts the request and returns a safe fallback response instead. When a policy calls for masking, Lasso returns the redacted text and the Gateway substitutes it before forwarding.
4. Optionally, once the model generates a response, the Gateway sends that output back to Lasso for a final scan, so the answer is checked for sensitive data or harmful content before it reaches the user.

Configuration is just as straightforward. You add Lasso as a guardrail in your TrueFoundry account, point it at your deployed endpoint, supply your API key, and attach it to the models you want to protect. From there it runs on every request automatically.

Get Started with Enterprise-Grade AI Security

AI security should not slow you down. With Lasso Security integrated into the TrueFoundry AI Gateway, you get purpose-built GenAI protection across every model and application, set up more like flipping a switch than running a complex engineering project.

To learn more, visit the Lasso Security integration reference in the TrueFoundry docs and the Lasso Security platform. Ready to see it in action? Reach out to either team for a demo and start securing your AI applications today.

‍