The recent launch of DeepSeek's advanced artificial intelligence (AI) models has sent ripples through the technology sector, prompting a reassessment of valuations among private AI companies. DeepSeek, a Chinese AI startup, has introduced low-cost AI models that challenge the dominance of established AI leaders, leading to late-January declines in the value of public tech stocks globally, followed by some tempered rebounds over the last few days.
In the private market, the impact is still evolving, but this development has led to a recalibration of valuations for AI and AI-related startups that is unfolding in real-time. Investors have become more cautious, likely favoring companies that demonstrate cost-effective innovation and efficient resource utilization. However, this does not necessarily mean that private market leaders, such as OpenAI, Anthropic, Mistral AI and Cohere, will suddenly face grave jeopardy. They remain some of the most celebrated AI trailblazers with respected models and they have already demonstrated their agility in the past.
Also, the market shockwaves we observed over the past few days could ultimately yield opportunities for private AI companies that begin to focus on lean(er) development strategies, leading to increased funding opportunities. In fact, in an exclusive January 30, 2025 article, the Wall Street Journal reported that OpenAI is in talks for a huge investment round led by SoftBank, which would potentially value the company up to $300 billion. Further, as VentureBeat noted on January 16, 2025, one of the key trends worth watching in 2025 is the continued drop in the cost of using AI models. Essentially, the lower cost of AI workloads potentially puts most AI services on sale, which could lead to more AI use and more applied-AI companies with better business models entering the market at competitive costs.
Here's why: since DeepSeek’s AI models are almost entirely open-source, large language model developers in the US and Europe will soon look to implement their efficiency gains in their own models. Therein lies even more opportunity: (i) a reduced need for future capital expenditures for US and European AI companies that are already on growth trajectories and (ii) the reduced need for large, costly computational resources. While this may not make public companies like NVIDIA very happy, it has given some emerging AI companies a better light at the end of the tunnel.
Another understated variable is the fact that some US-based enterprise companies may simply not trust a Chinese company with their AI workloads, after CNBC reported that President Trump noted that the sudden rise of the DeepSeek “should be a wake-up call” for America’s tech companies. This means that DeepSeek’s potential for penetrating the vast US market may be somewhat limited.
Of course, a basic analysis of what has transpired would be remiss if it did not mention the potential danger of overly inflated valuations in the current market. Nor is now the time for impulsive reactions, since the story continues to unfold and the fruits of data analysis at this juncture could be limited, while the story unfolds. Pundits need to learn a lot more about how this will impact private AI company valuations in the coming days before they can say with certainty that this is a serious ‘ongoing’ threat to private US and European AI companies. Nonetheless, the emergence of more cost-effective models like DeepSeek's will prompt investors to scrutinize valuations more closely, as they assess whether predominantly US-based AI companies can maintain their competitive edge against new, leaner entrants to the private market.
DeepSeek's launch events come at a pivotal moment in the AI industry, as they challenge existing paradigms and prompt a reevaluation of what constitutes ‘value’ in AI innovation. Private AI companies will need to adapt by emphasizing efficiency and cost-effectiveness to align with the evolving investor expectations. Further, the intensity of the market’s reaction to DeepSeek’s news is almost certainly a reflection of the significant expectations, high valuations and dominance of AI in the private market. While leading private AI companies, such as those mentioned above, may be healing from temporary bruises and pricing power questions, they are still well-poised to be some of the darlings of the AI sector.