Portal26 launches token controls for autonomous AI agents

Portal26 has launched an Agentic Token Control module for autonomous AI agents, aimed at helping businesses control spending as they deploy more agent-based AI systems.

The module gives organisations visibility into how many tokens their AI agents consume while running and lets them set limits in real time. It can monitor agents during operation, apply policy-based thresholds, and intervene when usage approaches or exceeds a defined budget.

The launch reflects growing concern among companies adopting agentic AI for complex workflows. Autonomous systems built on large language models can generate significant costs if they enter recursive loops, make repeated queries, or extend tasks beyond their original scope.

The controls can throttle, pause, or terminate an agent when token use exceeds set parameters. Customers also get a view of token consumption across agents, workflows, and broader organisational deployments.

Cost risk

Businesses are under pressure to move AI projects from pilots into everyday operations, while finance and technology teams try to avoid unpredictable spending. Token use has become a closely watched metric because it directly affects the cost of running language model-based applications and agents.

Portal26 positioned the issue as both a cost and operational risk. In addition to larger-than-expected bills, heavy or uncontrolled usage can affect system performance and create instability in automated processes.

"Agentic AI is powerful, but without cost controls, it can quickly become expensive and chaotic," said Arti Raman, Chief Executive Officer of Portal26.

"We've watched enterprises like Uber discover the hard way that adoption speed and cost predictability are on a collision course. Agentic Token Control gives organizations the telemetry and confidence to scale AI agents without waking up to an invoice they didn't plan for," Raman said.

The module is available immediately through Portal26's broader platform for managing enterprise adoption of generative and agentic AI. The company has recently added other tools designed to help customers oversee agent-based systems and measure business returns.

Governance layer

The release of token controls points to a wider shift in the AI software market, as suppliers move beyond model access and application building into operational governance. As companies deploy more autonomous tools across departments, they are increasingly looking for ways to track usage, set limits, and demonstrate that systems remain under human and policy oversight.

This is particularly relevant in sectors with stricter compliance demands, such as finance, insurance, and healthcare, where companies may need audit trails and formal controls around how AI systems are used. Portal26's customer base includes Fortune 500 companies, utilities, and regulated industries.

Pakshi Rajan, Chief Product and AI Officer at Portal26, described token management as a basic requirement for larger-scale use of autonomous AI. "Agentic cost controls represent a foundational layer for responsible AI operations," Rajan said.

"It's more than cost controls-it's about making agentic systems reliable, governable, and enterprise-ready," Rajan said.

Portal26 said no dedicated product had previously addressed agentic token risk at scale, as businesses work to keep AI systems within budget while expanding their use.