Open-source AI models are gaining mainstream use as enterprises, governments, and startups adopt them for flexibility and lower costs, while publishers and platforms debate how value is captured. Several sources also stress that “open” often refers to openly available model weights, not full openness around training data, recipes, or reproducibility.

TechCrunch frames the shift around questions of monetisation—how companies profit when models become widely available—and highlights trade-offs around cost, control, and risk. Dev.to adds a more operational angle, arguing that money increasingly moves to the surrounding “friction” such as hosted inference, managed deployment, fine-tuning, governance, and consulting, rather than the free model download itself. It also notes that licensing terms can affect real-world eligibility and that regulatory scrutiny can still apply despite permissive licences. Across coverage, Chinese open models are portrayed as gaining momentum, raising questions about whether American players can keep pace. The differing focus is largely between business strategy (who profits) and implementation and legal/regulatory considerations (what’s actually open and what costs remain).