DeepSeek Prioritizes AGI Over Profit, Founder Says Top AI Models Will Likely Stay Open Source

DeepSeek founder Liang Wenfeng said the Chinese AI startup is prioritizing the pursuit of artificial general intelligence (AGI) over maximizing profit and is likely to keep its most advanced AI models open source, outlining the company's long-term strategy during a recent meeting with investors.

According to a Yicai report cited by Reuters, Liang told investors that DeepSeek views open-source development and commercial monetization as compatible goals rather than competing priorities. His comments came during a four-hour meeting in which he discussed the company's research direction, computing needs and future infrastructure plans.

Liang said DeepSeek is likely to continue releasing its strongest models as open source, positioning the company differently from many leading AI developers that restrict access to their most advanced systems through proprietary platforms.

The startup gained international attention after its low-cost V3 and R1 models challenged assumptions about China's ability to compete with U.S. AI companies despite U.S. restrictions on exports of advanced AI chips.

DeepSeek recently completed its first fundraising round at a valuation of roughly $52 billion, marking a shift from its previous strategy of declining outside investment. The funding comes as competition intensifies from Chinese technology companies including ByteDance as well as AI startups such as Moonshot and Z.ai.

Liang also identified computing capacity as the company's biggest obstacle, saying the primary gap between Chinese and U.S. AI developers lies in access to computing resources rather than technical expertise.

Looking ahead, he said DeepSeek intends to build its own large AI computing clusters but has not decided whether it will develop proprietary AI chips. “I hope to be able to buy chips at a reasonable price so that I don't have to make chips myself,” Liang said, according to the Yicai report.

The comments come as U.S. export controls continue to limit China's access to some of the most advanced AI chips. Reuters previously reported that the United States has not added DeepSeek, Chinese memory chipmaker CXMT and more than 100 other companies identified as national security risks to a trade blacklist, citing efforts to avoid escalating tensions with Beijing.

Liang's remarks provide one of the clearest public indications of how DeepSeek intends to balance frontier AI research, open-source model development and commercialization as it expands its infrastructure and competes with both domestic and international AI developers.

This analysis is based on reporting from money.usnews.com.

Image courtesy of AI Supremacy.

This article was generated with AI assistance and reviewed for accuracy and quality.