Full Report
Part 4: AI ratchets up the throughput for telecom
Analysis Summary
# Industry News: Telecom Shifts to "Repatriated" IAM to Handle AI-Driven Throughput
## Summary
As AI and machine-to-machine automation scale within telecommunications, traditional SaaS-based Identity and Access Management (IAM) is hitting performance and cost ceilings. The industry is seeing a strategic shift toward "repatriating" core identity functions—bringing them back under direct operator control—to ensure network stability, low-latency authorization, and predictable cost structures.
## Key Details
- **Date:** March 17, 2026
- **Companies Involved:** Broadcom (via IMS Division), Telecom Operators (Global)
- **Category:** Market Trend / Strategic Analysis (Infrastructure Repatriation)
## The Story
The telecommunications sector is increasingly moving away from the "SaaS-first" IAM model for its core network operations. The rise of AI Ops, cloud-native functions (CNFs), and self-healing networks has transformed IAM from a "front door" for employees into a high-speed "control plane" for machines.
The industry is finding that public cloud IAM providers often struggle with the unique demands of telco infrastructure. When AI-driven automation triggers high-frequency cascades of authorization requests across distributed systems, SaaS-based identity services often experience rate-limiting, latency spikes, and unpredictable volume-based pricing. Consequently, telcos are reclaiming (repatriating) token services, machine identity management, and telemetry pipelines to ensure that identity functions operate at the same speed and reliability as the network itself.
## Business Impact
### For the Companies Involved
- **Broadcom/Infrastructure Providers:** Positioning themselves as the alternative to SaaS-only IAM, offering tools that support localized, high-performance identity control.
- **Telecom Operators:** Gaining greater operational resilience and "determinism," ensuring that identity outages do not bleed into network outages.
### For Competitors
- **SaaS IAM Vendors (e.g., Okta, Microsoft):** Face pressure to offer more flexible "edge" deployments or localized instances to prevent customers from pulling workloads back on-premises or into private clouds.
### For Customers
- **Enterprises/MVNOs:** Benefit from more stable and responsive telecom APIs, as lower-latency identity checks improve overall service performance.
### For the Market
- **The "Cloud Repatriation" Trend:** This highlights a broader market realization that the "all-in-on-SaaS" model has limits, particularly for mission-critical industrial and infrastructure sectors.
## Technical Implications
- **High-Throughput Authorization:** Shift from "authentication" (logging in) to continuous "transaction processing" for machine tokens.
- **Full-Fidelity Telemetry:** On-premises or private-cloud IAM allows for 100% log retention without the prohibitive costs of SaaS ingestion fees, essential for forensic chain-of-custody.
- **Latency Optimization:** Moving identity decisioning closer to the network edge (where CNFs operate) to avoid "trombone" routing to public cloud regions.
## Strategic Analysis
- **Market Positioning:** Identity is being redefined as "Critical Infrastructure" rather than just a "Security Service."
- **Competitive Advantage:** Operators who repatriate IAM can scale AI Ops more aggressively without incurring linear increases in identity licensing costs.
- **Challenges:** The primary risk is the increased operational overhead of managing complex IAM infrastructure in-house, requiring high-tier security engineering talent.
## Industry Reactions
- **Expert Commentary:** Analyst Jason Wilcox (Broadcom) emphasizes that telecom identity isn't just "enterprise IAM with better uptime"; it is a functional component of network stability.
- **Market Response:** Growing skepticism toward volume-based pricing models for machine-to-machine identities, which are seen as a tax on innovation.
## Future Outlook
- **Predictable Costs:** Expect a shift toward flat-rate or infrastructure-based pricing for identity services as volume-based SaaS models become untenable for machine-heavy telcos.
- **Hybrid Identity Models:** A "bifurcation" of IAM where SaaS is used for standard employee SSO, but private, repatriated systems handle the high-speed automated control plane.
## For Security Professionals
Cybersecurity practitioners in the telco and infrastructure space should evaluate their current IAM rate limits and latency. If AI automation is being "throttled" by your identity provider, it creates a "Catch-22" where responders cannot access systems during a crisis. Professionals should prioritize **deterministic recovery**—ensuring the identity system works even when the external internet or SaaS provider is unreachable.