Companies are taking a more hands-on approach to artificial intelligence (AI), creating their own ‘factories’ that harness data and tailor insights to specific needs. This move is driven by the desire for both scale and sovereignty—ensuring that critical information flows in a controlled, secure manner.

The challenge? Balancing this ownership with the need for high-quality data that can power reliable AI insights. At MIT Technology Review’s EmTech AI conference, speakers like Hewlett Packard Enterprise’s Chris Davidson and Oak Ridge National Laboratory’s Arjun Shankar discuss how governments and enterprises are rethinking AI to be more self-reliant.

Davidson leads efforts in building secure, scalable national- and enterprise-grade AI capabilities. Meanwhile, Shankar focuses on the intersection of computer science and scientific discovery, driving advancements in computational science and data models.

The conversation also touches on broader themes, such as OpenAI’s ambitious plans to fully automate research, or how games like Pokémon Go are inadvertently creating detailed world maps for AI training. Understanding these developments is crucial as we navigate the evolving landscape of AI technology and its implications for society.