DeepSeek: What it means for AI startups

DeepSeek’s entry into the large language model (LLM) market has sent shockwaves through the AI industry, challenging the dominance of tech giants like OpenAI, Anthropic, and Google DeepMind.

Its early success indicates that innovation in AI is not solely dependent on massive capital investments but can also thrive on open-source collaboration and iterative improvements. This development carries significant implications for startups, big AI investors, and venture capitalists (VCs) reconsidering their approach to AI funding.

For AI startups like Flare, which focus on building niche applications atop existing models, DeepSeek’s rise represents a major opportunity. “The prevailing notion that only a handful of companies could afford to compete in AI is now being questioned,” says co-founder and CEO, Magda Woods. “Startups can now build innovative solutions, leveraging open-source advancements rather than investing billions into training their own foundational models.”

The potential result? A more diverse AI ecosystem where problem-specific solutions flourish, rather than a monolithic race to build the biggest model.

Additionally, DeepSeek’s emergence could trigger a redirection of capital toward AI innovation rather than consolidation within a few dominant players. This challenges Sam Altman’s assertion that AI startups with only $10 million were “totally hopeless.” DeepSeek is proof that breakthroughs can emerge outside of well-funded Silicon Valley giants, opening the door for more startups to challenge incumbents.

DeepSeek’s success puts pressure on major AI companies to rethink their strategies. While OpenAI, Google, and Meta have relied on proprietary models as competitive moats, open research initiatives like DeepSeek threaten to erode this advantage. Investors in AI powerhouses—such as Microsoft (OpenAI’s backer) and Nvidia (which profits from AI hardware demand)—may need to reassess whether their bets on closed-source AI will continue yielding dominance.

Moreover, OpenAI’s reported accusations of intellectual property theft suggest growing tensions within the AI space. Critics argue that OpenAI itself built its models on vast amounts of publicly available internet data, making its stance on IP disputes somewhat tenuous.

For VCs, DeepSeek’s rise is an encouraging sign that smaller AI startups are viable investment opportunities. In Europe and the UK, where access to capital is more constrained than in Silicon Valley, DeepSeek’s model demonstrates that talent and innovation—not just deep pockets—can drive success. Peter Kane recently highlighted this in a NewsAgent interview, emphasising that AI breakthroughs don’t require billions, just the right approach.

While security risks remain a concern, DeepSeek’s contribution is less about application and more about proving what is possible. The AI industry is at an inflection point, and investors must decide whether to double down on Big Tech or embrace the disruptive potential of open AI ecosystems.