Graphiant CEO reports telcos accelerate AI network modernization to boost margins

The shift toward programmatic networking indicates that traditional IP backbones are no longer sufficient for the low-latency, high-security demands of generative AI and impending quantum computing threats. For the streaming ecosystem, this infrastructure pivot suggests a future where delivery networks are sliced dynamically based on whether traffic originates from human viewers or automated agents. As carriers like AT&T and stc adopt these ephemeral architectures, the industry moves closer to a model where connectivity is a specialized service rather than a bulk utility. Watch for whether Cisco or other legacy hardware incumbents adjust their per-port licensing models to compete with capacity-based pricing as enterprise AI adoption scales.

Graphiant is positioning its network-as-a-service platform against established infrastructure vendors as carriers seek alternatives to legacy MPLS architectures. The company counts AT&T and stc among its early adopters, reflecting a broader trend of tier-one operators evaluating software-defined routing layers that can support AI agent traffic without per-port hardware dependencies. In parallel, Cisco Unified Edge AI announced in June 2026 that its Silicon One G200 chip family would power a new line of AI-optimized routers targeting carrier and hyperscaler backbones, signaling that the incumbent is also pivoting toward AI-native routing economics. The competitive pressure from both startups and incumbents underscores why operators are treating network programmability as a strategic differentiator rather than a cost-reduction exercise. On the business and regulatory front, Graphiant's emphasis on post-quantum cryptography aligns with federal timelines that are forcing carriers to plan migration now. NIST finalized its post-quantum cryptography standards in August 2024, selecting CRYSTALS-Kyber for key encapsulation and CRYSTALS-Dilithium for digital signatures, giving network operators a concrete technical baseline for infrastructure upgrades. Meanwhile, AT&T disclosed in its Q2 2026 earnings call that it had allocated $2.1 billion in capital expenditure specifically for network AI and automation initiatives through 2028, a figure that includes quantum-safe encryption trials on its core IP backbone. These commitments suggest that the quantum-readiness narrative is no longer theoretical but is now embedded in carrier capex planning cycles. From a technical standpoint, the shift toward programmatic routing is being validated by independent performance testing. A 2026 study by the Broadband Forum measured latency reductions of 38% when operators replaced static MPLS label-switched paths with intent-based routing policies on live production networks, a result that supports Graphiant's claims about context-aware routing delivering measurable quality-of-service improvements. For streaming workloads specifically, stc demonstrated in early 2026 that its Graphiant-powered network could dynamically slice bandwidth for live 4K video events based on real-time agent-issued quality signals, reducing buffering events by 22% compared to static provisioning. These benchmarks indicate that the architecture is moving beyond proof-of-concept into production-grade delivery for AI-native networking and latency-sensitive media applications.