
TrustedRouter raises $6 million to compete on confidential AI routing
The AMW Read
The seed round adds a privacy-focused entrant to model-routing infrastructure, but its limited launch history and unproven adoption make this a subsegment update.
TrustedRouter raises $6 million to compete on confidential AI routing
TrustedRouter closed a $6 million seed round led by Sierra Ventures, the startup told Semafor, six weeks after its public launch. Led by CEO Joseph Perla, the company helps developers switch between AI models to reduce costs and dependence on a single provider. Its proposed distinction is end-to-end encryption and verifiable confidential computing for some open-weight models. Perla says those protections prevent both TrustedRouter and the hosting data center from reading customer prompts or outputs.
The financing adds a privacy-focused competitor to the AI infrastructure layer connecting applications with model providers. Routing makes model access easier to substitute, putting pressure on intermediaries to offer differentiation beyond a shared interface. TrustedRouter is betting that confidential execution can provide that distinction. But its addressable workloads depend partly on whether supported open-weight models meet customers' quality requirements. Semafor identifies that tradeoff as a constraint: lower prices and stronger confidentiality may not outweigh weaker outputs on demanding tasks.
For builders, the concrete evaluation is whether a supported model can complete a specific workload reliably under the promised privacy protections. Confidentiality coverage should be checked model by model, since the article describes it for only some open-weight offerings. Investors should distinguish this operating proposition from Perla's longer-term ambition for a token exchange where hosting providers compete for requests; the article presents that marketplace as an aspiration, not an available product.
#TrustedRouter #AIInfrastructure #ModelRouting #ConfidentialComputing #SeedFunding




