Open AI Models Gain Ground as CIOs Seek Lower Costs and More Control
Updated
Updated · Computerworld · May 19
Open AI Models Gain Ground as CIOs Seek Lower Costs and More Control
3 articles · Updated · Computerworld · May 19
Open-weight AI models are winning more enterprise use as IT leaders customize systems, keep deployments on-premise and reduce dependence on proprietary providers.
Free-to-download models such as Meta’s Llama, Mistral, DeepSeek and Minimax appeal because they offer more visibility into governance, lower computing costs and fallback options after outages at closed-model vendors.
Google, OpenAI and Microsoft have also released lighter open variants—Gemma, GPT-OSS and Phi—but analysts said these models are typically trained on less data and need more experimentation to find the right use case.
ServiceNow, Microsoft, HubSpot, RWS and Nvidia say open models fit agentic AI, infrastructure integration and robotics, while France and the UAE are using them to advance sovereign AI strategies.
Security remains a constraint: a UK-backed study warned open-weight models can be abused for attacks, and fixes cannot be pushed universally once users deploy their own versions.
Is the 'free' cost of open-source AI an illusion that hides the massive expense of specialized talent and security?
As companies flee vendor lock-in, are they walking into a minefield of unpatchable AI security threats?
Will the global race for 'Sovereign AI' build digital independence or a fractured world of warring AI ecosystems?
2026 Enterprise AI Shift: Anthropic Surpasses OpenAI as Cost, Trust, and Integration Challenges Mount
Overview
In April 2026, Anthropic's Claude overtook OpenAI's ChatGPT in adoption among American businesses, marking a major shift in the enterprise AI landscape. This change, highlighted by the Ramp AI Index, showed Anthropic's usage rising to 34.4% while OpenAI dropped to 32.3%. The Ramp AI Index, which tracks spending from over 50,000 U.S. businesses, revealed this turning point but also noted that Anthropic’s lead is not yet secure. The report emphasizes that while Anthropic gained ground, the AI market remains highly competitive and dynamic, with ongoing innovation and no clear long-term leader yet established.