Businesses turn to generative and agentic AI to cut costs

Businesses are combining generative AI with agentic systems to automate planning and execution across customer support, finance, hiring and software development. Generative AI produces content such as text, graphics and code, while agentic AI plans tasks, makes decisions and uses tools and APIs to carry out workflows. Together, these capabilities can reduce manual work and reliance on external software vendors.

Klarna disclosed that its OpenAI-powered customer-facing assistant handles almost 80% of support calls, generating annualized savings of about $59 million. IBM reports a 40% absolute reduction in administrative costs related to human resources through localized AI-driven automation. Corporate spending priorities also favor systems that offer measurable consumption efficiency and localized cost control, according to a Gartner Executive Report.

Practical applications include autonomous exception handling, workflow redesign and internal software development. Businesses can use Microsoft Copilot, Google Gemini and ChatGPT for general productivity; Perplexity AI and Fireflies.ai for research; and Zapier and Notion AI for workflow automation. Anthropic Claude, Cursor and GitHub Copilot support writing and coding. The central opportunity is to integrate these capabilities into core business processes, rather than limit their use to standalone content generation.