In the rapidly evolving world of artificial intelligence, two companies have recently made headlines: OpenAI and DeepSeek.
OpenAI, a U.S. based leader in generative AI, has developed some of the most widely used AI models, including ChatGPT and GPT-4, which have revolutionized natural language processing.
On the other hand, DeepSeek, a relatively new Chinese AI startup founded in 2023, has quickly emerged as a formidable competitor by developing open-source large language models that rival those of more established firms.
The competition between these two companies is becoming a significant AI industry rivalry, which has called America’s AI-dominance into question.
DeepSeek’s Rapid Rise in the AI Landscape
DeepSeek has positioned itself as a major player in AI by focusing on cost-effective, high-performance models whose performance challenge those of industry giants like OpenAI and Google DeepMind. The company has received substantial backing from investors, particularly in China, where AI development has been accelerating rapidly.
A key aspect of DeepSeek’s strategy is the use of “distillation”, a technique that trains smaller models to replicate the performance of larger, more complex AI systems. This enables DeepSeek to develop scalable models more efficiently, reducing the cost of deployment while maintaining strong performance. DeepSeek has also made several of its models open-source, attracting developers who prefer AI tools that are not locked behind proprietary paywalls.
DeepSeek’s latest model, DeepSeek-Coder, is a direct competitor to OpenAI’s Codex and Google’s Gemini AI, providing powerful code-generation capabilities. The release of such models has positioned DeepSeek as a growing threat to OpenAI’s market share in the AI ecosystem.
The dispute between OpenAI and DeepSeek
The rapid ascent of DeepSeek has divided the AI community. Some developers and AI researchers believe that DeepSeek’s open-source approach represents a positive step toward democratizing AI, ensuring that powerful tools are not monopolized by a few Western tech giants. Others, however, argue that DeepSeek’s reliance on OpenAI’s research and methods crosses ethical and legal boundaries.
A dispute between OpenAI and DeepSeek has escalated for two main reasons.
1. Intellectual Property Concerns
OpenAI has accused DeepSeek of reverse-engineering its proprietary AI models. This raises legal and ethical questions about the boundaries between legitimate and illegitimate AI development and whether DeepSeek has benefited from OpenAI’s prior research in a way that breaches intellectual property laws.
2. Data Scraping Allegations
Microsoft and OpenAI are investigating whether DeepSeek gained access to its proprietary data to train its own AI model. There are also claims that DeepSeek trained its models using OpenAI-generated text responses. Although the text responses are publicly available, OpenAI executives argue that this practice undermines fair competition and poses potential security risks and intellectual property violations where sensitive or copyrighted content is involved.
Implications of DeepSeek’s Rapid Ascent
DeepSeek’s rapid growth and adoption by users in China and worldwide has a number of significant implications.
1. Changing the Structure of the AI Industry
Tech investors are closely monitoring the situation, as DeepSeek’s low cost AI-development approach could reshape the AI industry.
Nvidia’s stock price has already taken a beating, as equity analysts debate whether DeepSeek’s low cost approach to training AI models will obviate the need for large and expensive data centres, which demand large numbers of Nvidia chips.
2. Calls for Increased Regulation
The U.S. government and NGOs have expressed concerns about China’s rapid advancements in AI, which could enable firms like DeepSeek to circumvent restrictions on AI model training and compute power that are designed to ensure AI safety. Some experts have called for new AI regulations due to fears that lack of oversight and mutual cooperation could lead to an arms race in AI development. This could slow innovation, but at the same time ensure greater transparency in AI research.
3. More Open-Source AI Projects
DeepSeek’s rise demonstrates the viability of open-source AI models. Thus, more firms are likely to invest in new or existing open-source models, which allow them to compete head-to-head with more expensive proprietary systems from companies like OpenAI, Google, and Anthropic.
4. Potential for New Partnerships
Despite their rivalry, there is also potential for OpenAI and DeepSeek to explore collaboration opportunities, particularly if international AI regulations require companies to share safety research to ensure responsible AI development.
The bottom line
The rivalry between OpenAI and DeepSeek reflects the broader challenge facing the AI industry — balancing rapid AI advancement with the need to ensure that AI development remains beneficial to society.
Whether the current rivalry leads to stricter regulations, new business models, or even unexpected partnerships, it will likely shape the next round of AI development worldwide.
Zuhair Imaduddin is a Senior Product Manager at Wells Fargo. He previously worked at JPMorgan Chase and graduated from Cornell University.
Image: DALL-E
Follow us now on LinkedIn.