The artificial intelligence (AI) industry is witnessing a concerning trend, where major players are using misleading marketing tactics to promote their AI models as "open" and "transparent." Companies like OpenAI, Google, and Microsoft are employing a strategy known as "open-washing," which involves overstating their commitment to openness while keeping critical components proprietary. This approach is not only misleading but also perpetuates the concentration of power in the hands of a select few.

The term "open-washing" is borrowed from the concept of "greenwashing," where companies exaggerate their environmental credentials. In the context of AI, open-washing refers to companies presenting certain parts of their pipelines as open while controlling much of the ecosystem and extracting value from users seeking to customize or extend their tools. This strategy is particularly prevalent in cloud computing, where many of these open-washed models exist on cloud providers and are built and sold by cloud providers as well.

A closer examination of these "open" AI models reveals that critical aspects, such as data sets, infrastructures, training methods, and even the practical use of large language models (LLMs), remain tightly guarded. Companies maintain intellectual property dominance by presenting certain parts of their pipelines as open while controlling much of the ecosystem and extracting value from users seeking to customize or extend their tools. For instance, the release of LLMs "under permissive licenses" that claim anyone can use or adapt them may appear to democratize AI, but these models often exclude access to critical features such as complete training data sets or the computational power needed to replicate the models from scratch.

The distortion of the principles of openness, transparency, and reusability lies at the heart of open-washing. Transparency in AI would entail publicly documenting how models are developed, trained, fine-tuned, and deployed, including full access to the data sets, weights, architectures, and decision-making processes involved in the models' construction. Most AI companies fall short of this level of transparency, selectively releasing parts of their models and crafting an illusion of openness. Reusability, another pillar of openness, is also compromised, as companies allow access to their models via APIs or lightweight downloadable versions but prevent meaningful adaptation by tying usage to proprietary ecosystems.

The development of generative AI hinges on immense resources, including massive data sets, computing power, and specialized frameworks. Training state-of-the-art LLMs requires computational energy and hardware resources, making it inaccessible to most enterprises. This concentration of power in the hands of a few corporations raises concerns about the democratization of AI. Even permissively licensed models, such as Meta's Llama 3, come with restrictive terms limiting how they can be deployed or adapted, ensuring smaller organizations remain dependent on these corporations' ecosystems.

Enterprise leaders need to be cautious when evaluating AI models marketed as "open." It is essential to ask tough questions, such as what exactly can be modified, where is the complete documentation, and can the model be taken and run with it anywhere desired. The restrictions and limitations imposed by these companies may not align with the expectations of openness and transparency. As the AI landscape continues to evolve, it is crucial for enterprises to be aware of these misleading marketing tactics and focus on what these AI tools can do for their business within their constraints.

In conclusion, the trend of open-washing in AI is a concerning development that perpetuates the concentration of power in the hands of a select few. Enterprises must be vigilant and not fall prey to these misleading marketing tactics. By understanding the limitations and restrictions imposed by these companies, businesses can make informed decisions about their AI strategies and avoid getting caught up in the hype. As the AI industry continues to grow, it is essential to promote genuine openness, transparency, and reusability, rather than relying on performative marketing strategies.