Every one of Cisco's 90,000 employees now has an AI agent. Its CFO explains the cost

Cisco is preparing to give every one of its approximately 90,000 employees a personalised AI agent as part of a company-wide rollout beginning in its new fiscal year at the end of July. While the initiative marks one of the largest enterprise AI deployments to date, the company says managing the cost of AI is just as important as scaling it.

Speaking to Fortune, Cisco Chief Financial Officer Mark Patterson said the company has built its AI architecture to prioritise efficiency rather than relying on expensive frontier AI models for every task.

"We're not going to burn a whole bunch of tokens with frontier models," Patterson said, explaining how Cisco intends to control AI costs while expanding access across its workforce.

Efficiency, not bigger AI models, is the priority

Rather than assigning every request to the most advanced large language model, Cisco has built an AI system that dynamically selects the most appropriate model for each task.

According to Patterson, the platform determines which AI tool is both the most effective and the most efficient before processing a request.

The company has also developed much of its AI infrastructure on-premises, giving it greater control over both operating costs and enterprise data.

Cisco did not disclose the financial investment behind the rollout, noting that such costs are not reported separately outside its regular earnings disclosures.

Every employee gets a personalised AI assistant

Under the rollout, each employee will receive an AI agent capable of supporting day-to-day work by:

• Answering workplace questions
• Completing routine tasks
• Routing requests to the most suitable AI model
• Helping employees work more efficiently across functions

Unlike conventional AI chatbots, AI agents can independently plan tasks, call external tools and complete multi-step workflows. These capabilities also make them significantly more resource-intensive.

Industry estimates cited by Fortune suggest a typical chatbot interaction may consume only a few thousand tokens, while complex AI agent workflows can require hundreds of thousands, or even millions, of tokens during a single task. This makes efficient model selection increasingly important as organisations deploy AI at enterprise scale.

Finance is already seeing measurable gains

Cisco has already begun integrating AI into its finance operations.

Patterson said AI now produces 80% to 90% of the first draft of the company's Management Discussion and Analysis (MD&A), the narrative section included in public company financial filings.

The finance team has also developed an AI-powered investor relations tool that analyses Cisco's historical financial performance alongside competitors' earnings calls. The system is designed to anticipate questions from individual financial analysts ahead of earnings announcements.

Another initiative under development is a "CFO cockpit", an AI-powered dashboard intended to combine data across products, geographies and customer segments while generating forward-looking business insights and recommendations.

Training accompanies the technology rollout

Cisco is pairing the deployment with company-wide upskilling and knowledge-sharing programmes to encourage employees to experiment with AI in their daily work.

Patterson said he expects teams to discover new business applications as adoption increases across the organisation.

He also uses his own AI agent, primarily to benchmark Cisco's performance against competitors across measures such as revenue growth, earnings per share, research and development spending and capital allocation.

AI remains central to Cisco's growth strategy

Founded more than four decades ago as a networking company, Cisco has expanded its AI strategy beyond enterprise infrastructure into data centre networking, custom silicon, optical networking and AI security.

According to Patterson, demand from hyperscale customers continues to accelerate. Key business figures shared by the company include:

• Approximately 90,000 employees set to receive AI agents.
• $2 billion in AI-related orders reported in fiscal year 2025.
• $9 billion in AI order guidance for fiscal year 2026.
• Around 53% growth in Cisco's share price year to date in 2026, according to Fortune.

Patterson, who has spent 26 years at Cisco and became Chief Financial Officer in July 2025 after serving as Chief Strategy Officer, described AI as the biggest technology transition of his career.

For Cisco, the next phase of enterprise AI is no longer centred solely on building infrastructure. It is also about making AI practical, scalable and economically sustainable for an entire workforce.