How DeepSeek could shake up the UK’s AI industry

Chinese artificial intelligence (AI) startup DeepSeek has recently hit the headlines for its large language model (LLM) release, which is set to challenge the likes of OpenAI and other popular AI systems.

DeepSeek, which was founded in 2023, has surpassed tools like ChatGPT and other competitors to become the highest-rated free app on the Apple App Store in the UK, US and China.

But despite its fast success, DeepSeek has also faced criticisms and concerns, particularly around its models potentially reflecting China’s political views. Moreover, its cutting-edge technology could make it difficult for UK AI startups to compete.

What is DeepSeek?

Deepseek is an AI lab which develops open-source LLMs. It was founded in 2023 by Liang Wenfeng, who also founded the Chinese hedge fund High-Flyer. The company focuses on developing advanced AI models and has gained traction for its cost-effective and high-performance solutions.

Its advanced reasoning model, DeepSeek-R1, competes with OpenAI’s o1 in performance – excelling in tasks like mathematics, coding and problem-solving. The model and its variants, including DeepSeek-R1 Zero, use advanced training methods like reinforcement learning (RL) and multi-step processes to build their skills and improve performance on tasks.

What makes DeepSeek stand out?

DeepSeek has caused a stir in the AI space, mainly because it allegedly released groundbreaking models that deliver performance on par with the world’s top chatbots, but at a fraction of the cost. Researchers behind its model claim that it was developed for less than $6 million, significantly less than its competitors.

Its commitment to being open-source has also made it stand apart from rivals. Unlike its US counterparts, such as OpenAI and Meta, DeepSeek’s technology is available to everyone for free. This means developers worldwide can access and improve DeepSeek’s models, making it easier to collaborate and share ideas.

However, people have also raised concerns that because DeepSeek was developed in China, its system might be influenced by the country’s political views, especially when it comes to concerns regarding human rights. Critics worry that these models could be designed to avoid or downplay controversial issues in a way that aligns with the Chinese government’s stance.

What does this mean for UK AI startups?

While DeepSeek has garnered a lot of positive attention for its unique approach, it’s likely to be a challenge for UK AI startups for several reasons.

Most notably, DeepSeek’s fast advancements and global reach could easily overshadow smaller businesses. After all, its model has already been able to demonstrate cutting-edge performance in tasks like reasoning, coding and mathematics. As a result, UK startups may struggle to match these capabilities without significant investment in research and development.

Additionally, because its technology is open-source, it gives other businesses and developers access to high-level AI tools without the need to create their own from scratch. This makes it easier for larger companies to integrate DeepSeek’s models into their products and services, leaving smaller businesses at a disadvantage.

DeepSeek’s rapid rise is creating both opportunities and challenges for startups in the AI space, as its cost-effective and high-performance models could put a lot of pressure on competitors, especially smaller businesses and startups.

Still, there are plenty of AI startups that have disruptive potential, especially those featured in the Startups 100 Index for 2025. These companies, such as ROBIN AI and MAGIC AI, are pushing the boundaries of AI technology in unique ways, proving that there is still room for smaller companies in such a competitive market.