Businesses Embrace Open-Weight AI Amid Heavy Tech Costs

American companies are increasingly using lower-cost "open" artificial intelligence models to manage skyrocketing tech expenses, the Financial Times reported. Open-weight models, which users can operate on their own hardware, often prove more economical compared to paying for advanced models from companies like OpenAI and Anthropic, according to AlphaSense data cited by the FT.

Earnings calls and investor conferences saw a significant increase in references to "open weight" or "open source" models, with mentions jumping sixfold in August and September compared to the same period in 2025. While the trend is particularly evident among tech firms, businesses in various sectors such as PNC Financial Services, logistics company CH Robinson, and industrial giant Siemens have also explored using open-weight models recently.

For instance, Tinder's chief technology officer, Vinay Kuruvila, explained that the company had shifted some queries from non-technical users to open-weight models due to escalating AI costs, noting a transition from $1 million per year in January to $10 million by July. The report suggests that the battle between open and closed source AI is more financially driven than ideologically, with CFOs weighing the potential savings, flexibility, and control open models offer against the responsibility of managing their underlying infrastructure.

As AI moves beyond isolated chatbots and starts operating across finance, procurement, treasury, compliance, and enterprise software, the significance of this tradeoff is expected to grow.