Intelligence Community Leverages AI

Posted 1/22/24

(NewsUSA) - The future of the intelligence community will hinge on how well intelligence services can adopt and leverage artificial intelligence (AI) tools to make the most of the constant flow of …

This item is available in full to subscribers.

Please log in to continue

Log in

Intelligence Community Leverages AI

Posted

(NewsUSA) - The future of the intelligence community will hinge on how well intelligence services can adopt and leverage artificial intelligence (AI) tools to make the most of the constant flow of data from all parts of the world, according to experts at the Special Competitive Studies Project (SCSP), a bipartisan nonprofit organization.

The U.S. Intelligence Community (IC) should take immediate action to leverage emerging generative AI (GenAI) capabilities to protect the nation and maintain our competitive advantage over global rivals, according to a new report from SCSP. The IC has been at the cutting edge of adopting emerging technology for decades, but GenAI will have an exceptionally broad impact, and will further complicate the already complex world of intelligence.

“Intelligence services will eventually need to plan and account for AI-enabled machines acting as semi-independent actors, directing operations and making decisions, both for our adversaries and allies,” according to the SCSP report.

Large Language Models (LLMs), the multi-modal models that can generate images, video, and sound, and other forms of GenAI will become more numerous, faster, more accurate, and more capable, according to the report. IC agencies will need to adapt their approaches to all aspects of the intelligence cycle, from planning and collection to analysis and dissemination.

The rapid expansion of AI provides the IC with an opportunity to make the changes needed to succeed in the digital era; the SCSP offers four actions for the U.S. intelligence community to consider.

Start using AI tools now. Training the IC workforce will help the IC create the infrastructure and policies needed to manage AI.

Be an agile adopter. ICs should partner with a foundation model provide to acquire their large language model, rather thran creating one from scratch.

Take on privacy concerns. The IC should tackle privacy issues head-on, giving IC agencies the authorization they need to secure and use personal data if necessary.

Collaborate when possible. The nature of intelligence is for different agencies to safeguard their own unique datasets and capabilities. However, the IC will need to strike a balance and promote coordinated approaches to innovation. Collaboration will not only yield economies of scale, but will avoid the proliferation of smaller LLMs that will pale in comparison to those of global rivals, such as the People’s Republic of China.

The SCSP also advises the U.S. intelligence community to create a committee on AI deployment and use to advise the DNI on the creation of IC datasets and AI algorithms to ensure access and responsibility.

For more information, visit scsp.ai.