The emphasis is increasingly moving from addressing the digital divide to tackling the emerging generative AI divide. It's important to ensure everyone has equal access to and understanding of generative AI technologies.
In a recent magazine article, Larry Irving shared his experiences while leading the National Telecommunications and Information Administration (NTIA).
During his tenure, he and his colleagues identified what became known as the digital divide while researching the growing gap between those who had internet access and those who did not.
Irving hoped that the U.S. government would eventually allocate $42 billion to bridge this divide, and thirty years into the Internet age, those funds became available to ensure that every American could get online.
The term "digital divide" became so prominent that it was featured in an NTIA report titled "Falling Through the Net: Defining the Digital Divide" and was also mentioned in a 2000 Presidential State of the Union address. Fast forward to July 2025, and the focus has shifted to a new concern: the "GenAI Divide."
Fast forward to July 2025, and the focus has shifted to a new concern: the "GenAI Divide."
A report titled "The GenAI Divide: State of AI in Business 2025" indicates that despite $30-40 billion in enterprise spending on generative AI, 95% of organizations are seeing no business returns.
The report's authors noted a stark divide among different types of enterprises, such as mid-market, small, and medium-sized businesses, startups, vendors, and consultancies. They reported that only 5% of integrated AI pilots can generate millions in value, while the vast majority struggle without a measurable profit and loss impact.
The report notes that a small percentage of organizations are achieving significant value by adopting learning-capable, highly customized AI systems through strategic partnerships.
The report concludes that overcoming this divide requires a shift from internal building to external partnerships, empowering line managers, and selecting adaptable tools that integrate deeply, leading to the emergence of an "Agentic Web" of interconnected, learning AI systems.
Other observers dispute this characterization, arguing that while it's accurate that only 5% of custom enterprise AI tools make it to production, labeling the 95% as a failure rate oversimplifies the issue.
Challenges such as fragile workflows, poor contextual learning, and misalignment with operational needs are contributing factors.
Sukh Sandhu pointed out on social media that generative AI is just one aspect of the broader AI landscape.
Sukh Sandhu, a specialist in higher education, cybersecurity, and artificial intelligence, pointed out on social media that generative AI is just one aspect of the broader AI landscape, and many are overlooking other powerful technologies that are already influencing the future. Sandhu emphasized that organizations that will thrive in the future are those that prepare for the entire AI landscape now.
- For example, technologies that provide forecasting, analytics, and action recommendations are widely used.
- Generative/Creative AI tools like ChatGPT and DALL·E create text, images, and video are prevalent.
- Neuro-symbolic AI is an emerging technology that combines reasoning, logic, and learning for use in areas like legal and scientific applications.
- Adaptive AI continually learns in real-time, with applications in process control.
- Agentic AI, which are autonomous, goal-driven agents, such as AI nurses and customer service bots, are rapidly emerging.
- Multi-modal, general-purpose AI models that work across various media (text, audio, image, video) are in a growth phase.
- Federated AI is expanding decentralized learning focused on privacy.
- Causal AI or advanced prototypes that understand cause-and-effect relationships are still in the research stage.
- Neuromorphic and Quantum AI experimental technologies are inspired by brain mechanisms or use quantum computing.
- Speculative AI, concepts like Artificial General Intelligence (AGI) and self-aware AI, remain theoretical.