Rebalance Contracts and Controls for Surging Agentic AI Traffic

What happened

Decodo's analysis finds automated bots now generate the majority of web requests and agentic AI traffic surged sharply year-on-year. The most material detail is that agentic activity concentrated on commercial pages—product and search—so buyer-facing sites see the biggest exposure. Watch whether suppliers start limiting API access or offering bot tiering that shifts cost and capacity responsibility to customers

Why the category manager should care

Treat bot and agent traffic as an active driver of supplier obligations—demand telemetry, capacity obligations and defined mitigation responsibilities in contracts

Key facts

• Agentic AI traffic rose significantly year on year (Decodo analysis)
• Bots represented the largest share of analysed web requests
• 77% of agentic requests reached product and search pages

What happened

Gartner frames the new AI era around agentic systems, rising costs and widespread maturity gaps. The operationally important detail is that agentic agents can create high volumes of model requests and token consumption that many organisations don't yet track or control. Watch vendor pricing and finance pushback as organisations try to make consumption predictable

Why the category manager should care

Insist on model-usage reporting, pre-approval gates and cost-control features in supplier offerings to prevent unexpected consumption charges

Key facts

• Analysts highlight model efficiency and agent orchestration as primary vendor competition areas
• Only a minority of organisations are at high AI maturity and many are still in experimentation

What happened

Industry voices report a growing enterprise challenge: connecting, managing and governing many AI agents alongside human teams. The important operational detail is the need for orchestration and policy mechanisms that match tasks to appropriate models to control token use. Watch whether vendors ship agent-management gateways or organisations adapt existing API gateways for this purpose

Why the category manager should care

Plan for procurement to include agent onboarding, role-based access rules and supplier commitments for agent lifecycle management

Key facts

• Discussion of matching agent workloads to appropriate models to optimise cost and consistency
• Enterprises are moving from experimentation toward implementation planning for agentic systems

What happened

Visa has expanded AI-powered security, governance and human-validation practices as part of a resilience strategy. The concrete detail: Visa operates many internal AI models and pairs automation with human oversight and audit-ready controls, which acts as a commercial benchmark for suppliers. Watch for regional partners and vendors to adopt similar governance standards that buyers can reference in contracts

Why the category manager should care

Use enterprise buyers' governance playbook to demand human‑validation steps, audit trails and explainability from suppliers that handle sensitive or transactional workloads

Key facts

• Visa operates hundreds of AI models internally and pairs automation with human validation
• The company highlights large-scale investment in AI security and governance capabilities

What happened

SecurityBrief's expert-opinions section collects practitioner columns stressing operational maturity, observability and governance over adding more tools. The practical point is contributors urge firms to improve runbooks, telemetry and governance to avoid backlogs as AI accelerates. Watch which operational patterns practitioners recommend becoming requirements in sourcing templates

Why the category manager should care

Don't rely solely on supplier tools—contract for operational outcomes such as telemetry retention, runbook updates and measurable governance evidence

Key facts

• Series of expert columns focusing on AI, observability and operational maturity
• Multiple pieces emphasise that operational foundations matter more than tool count