Editor’s note: This article has been updated to correct a quote by Joe Schurman that misstated the estimated value of the space economy by 2040.

In the depths of space, a cargo ship containing specially packaged food approaches the International Space Station, carrying out a resupply mission.

Without direct human control, the cargo ship docks onto the space station, transferring the food to the crew onboard.

In another instance, a U.S. satellite autonomously adjusts its orbit to avoid colliding with debris.

This is the future for robots in space, leveraging the power of artificial intelligence, industry experts said Thursday.

“The future is going to be autonomous maneuvering in space,” said Alexandr Wang, founder and chief executive officer of Scale AI, during a discussion on generative AI on the final day of the 40th Space Symposium at The Broadmoor. Scale AI is a San Francisco-based data annotation company that trains frontier AI models like OpenAI and others using labeled data.

In the ever-increasingly crowded heavens, agentic AI will be a crucial tool to help the U.S. orchestrate its space assets and warfighting capabilities, adversarial assets, space situational awareness and space domain awareness, Wang said. Agentic AI capabilities manage space debris, help spacecraft navigate autonomously and enable a spacecraft to plan and execute space missions, among other tasks. Then, it can raise key courses of action to military commanders who can make strategic and tactical decisions.

“Agentic AI will help uplevel the decisions and the capabilities that humans have in these processes,” he said. “AI can do a lot of minutiae and a lot of lower level planning and integration. But then, ultimately, we’re going to need to elevate these strategic decisions to humans. They’re going to have to think through what are the diplomatic implications? What are the strategic implications? How do I integrate softer tacit knowledge or other pieces of information to be able to conduct the perfect course of action going forward?”

Pointing to DeepSeek, the Chinese technology startup that since 2024 has released two artificial intelligence large language models rivaling the programs built by American companies, and which were built much more inexpensively, Wang said it was “very concerning” and “a huge wakeup call” for the entire space industry, which believed China was one or two years behind leading Western labs.

In a rapidly moving innovation cycle and amid increasingly competitive relations, the U.S. must ensure its AI capabilities exceed China’s, he added.

“We have an imperative. We need to match that innovation in the U.S., and, if anything, demonstrate that we’re more out in front of it,” Wang said.

Artificial intelligence has been used in space since at least 2019, mostly to sift through and make sense of data from sensors. Companies and the military needed ways to bring large datasets from space, like satellite imagery or weather data, Joe Schurman told The Gazette in a separate interview. Schurman is a PricewaterhouseCoopers partner specializing in AI for aerospace and defense and co-leader of the company’s U.S. space practice.

The space industry built on its success using AI and expanding its technical capability between 2019 and 2023. Now, the industry also uses generative AI to streamline aerospace engineering, project management and operations needed to simplify engineering and analytics to facilitate more rapid launches that will enable the commercialization of space, he said.

“AI can streamline mundane tasks, specifically agent-based AI. Each engineer has their own competency and specialty, such as payloads, et cetera. They all have to come together to launch a mission. If you pair up an engineer with an (AI) agent or a LLM (large-language model) chat, it will power the individual as a force multiplier to complete the engineering quicker,” Schurman said. “If you had a flight test engineer (working with an AI agent), it could help digitize the flight test (data) and streamline their ability to get data to structural design and other parts of the process.”

Most major aerospace and defense companies embrace generative AI, but they are limited in how they can use AI with classified data, he said. Over the last year, though, some AI services have come online that can be used in classified environments.

“Using those would enable you to get your fleet and payload out more quickly and enable the commercialization of space,” Schurman said.

Schurman said wider use of AI in space could create new space-based capabilities and allow greater use of space for many different businesses.

“AI can reduce years of engineering time to months and sometimes weeks,” he said.

The agriculture, pharmaceuticals and financial industries could all benefit from using AI for space-based capabilities and setting up operations in space, he said.

“That is how you get to a $2 trillion space economy by 2040,” Schurman said, as predicted in Space Foundation’s most recent report. “That will facilitate space industrialization.”

He said President Donald Trump will soon sign an executive order that will reduce the number of Federal Acquisition Regulations (FAR) needed to get contracts from the Space Force and the other military branches, which would make it easer for small businesses to do business with the military.

“That will mean more jobs, opportunities and further commercialization of space. There will be more contracts based on commercial instead of government terms. That will put startups on the same footing as major prime contractors from a contractual perspective,” he said.

Commercial companies are now 80% of the space economy, which totaled $570 billion in the Space Report issued this week. Space Foundation predicts that will grow to $300 trillion by 2040.

“This is a big deal here (Colorado Springs) with all the startups and operations you have here. The ability to grow is significant,” Schurman said.