DARPA director Steven Walker made the announcement at the agency’s 60th-anniversary conference.
“We want to explore how machines can acquire human-like communication and reasoning capabilities, with the ability to recognize new situations and environments and adapt to them,” Walker said.
The agency said that the money would fund projects on top of the 20 active programs exploring applications of AI, such as in cybersecurity or in human-computer symbiosis programs that examine the interaction between people and machines.
WATCH: Is Artificial Intelligence really more dangerous than a nuke?
The announcement comes as tech companies are actively researching AI as well, with companies such as Google impressing with new feats, such as an AI program defeating the world’s best player at the complex game Go in 2017.
Other governments, such as Canada, China, India and France, are prioritizing AI and have claimed it is key to growing their economies. China has said it wants to be a global leader by 2030.
Tesla founder and CEO Elon Musk has warned that AI will be a more dangerous weapon than the atomic bomb at SXSW in March.