Atsign and Intel claim 88x encryption uplift for edge AI agents
Atsign and Intel have jointly published a Smart Transportation Solution Brief claiming hardware-accelerated, end-to-end encrypted communications between autonomous AI agents run up to 88 times faster when Atsign's identity-first platform runs on Intel silicon. The companies say the result removes what they describe as a longstanding performance bottleneck that has limited practical deployments of agentic AI at the edge.
According to the published benchmarks, the performance uplift varies by processor. Running on Intel Xeon hardware, the combination achieved roughly 88x faster encrypted agent-to-agent throughput, reaching up to 5 Gbps. On Intel Core Ultra 9, the figure was approximately 60x, with throughput of up to 6 Gbps. Atsign did not disclose the baseline configuration used for comparison, nor the test conditions or independent verification of the figures.
The technical combination
Atsign contributes identity management, namespace isolation, scoped data sharing and auditability to the joint architecture. Intel contributes hardware-backed key generation, accelerated encryption and hashing, protected key storage, secure boot, Trusted Compute, Total Memory Encryption and Full Disk Encryption. The companies say the result is a zero-trust communications layer suited to distributed transport infrastructure, covering intersections, traffic corridors and mobility networks where AI agents must make safety-critical decisions without constant central oversight.
Intel has also admitted Atsign to its Industrial Builders partner programme, which gives Atsign pre-validated status on the Intel ecosystem directory. For Intel's customer base, that designation is intended to simplify procurement, reduce integration risk and provide access to Intel technical support for complex deployments.
Aparna Rayasam, chief executive of Atsign, said: "Intel and Atsign have delivered a genuinely secure foundation for agentic edge AI deployments in smart transportation environments, which are no longer constrained by something as mundane as technical performance."
Market context and competitive landscape
Securing agent-to-agent communications at the edge is an emerging design challenge as agentic AI frameworks move from data-centre experimentation into operational infrastructure. Smart transport is one of the first sectors where the stakes are high enough to make encryption latency a genuine deployment barrier: a traffic-management agent waiting milliseconds for a cryptographic handshake before issuing a signal change creates unacceptable risk at scale.
The zero-trust edge security space includes established players such as Zscaler and Palo Alto Networks on the network side, alongside specialist IoT identity vendors. Atsign's differentiation, based on identity-first architecture with no exposed inbound ports and no reliance on static IP addresses, positions the company in a narrower niche focused on device and agent identity rather than perimeter control. Admission to Intel's partner programme adds commercial validation and should widen distribution through Intel's reseller and systems-integrator network.
Regulatory backdrop
Smart transport deployments in Europe will need to satisfy the requirements of the NIS2 Directive, which extends cybersecurity obligations to transport operators and critical infrastructure providers from late 2024 onward. In the United States, the Department of Transportation's Connected Vehicle programme and emerging guidance from CISA on AI in critical infrastructure both push toward encrypted, authenticated machine-to-machine communications. Neither framework specifies hardware-accelerated encryption, but the performance gains Atsign and Intel are claiming address a practical concern regulators have implicitly acknowledged: that mandatory security controls must not degrade real-time operational performance.
No named transport operator customers are cited in the release. The next meaningful milestone for the partnership will be a named production deployment or an independently audited benchmark result.