Here’s what’s on our plate today:

• 🧐 DeepSeek’s R1 shook markets—and why global bans now threaten its R2.

• ❓ Test your guess on the real roadblock holding back DeepSeek-R2.

• 💡 Summarize the R1 upheaval (≤150 words) and give one tip for AI investors.

• 🚀 Cloudflare wants bots to pay, and AI diagnoses better than doctors.

Let’s dive in. No floaties needed…

The Laboratory

Beyond the initial boom, what is going on with DeepSeek

Few organizations can disrupt their industries the way DeepSeek managed with its R1 large language model. Within days of its release, the DeepSeek AI assistant, a mobile app that provides a chatbot interface for DeepSeek-R1, hit the top of Apple’s App Store, outranking OpenAI’s ChatGPT mobile app. But this was just the beginning. The R1 claimed to have been developed at a fraction of the cost when compared to models with similar capabilities by other U.S.-based companies. This information and the meteoric rise of DeepSeek in terms of usage and popularity triggered a stock market sell-off, as investors cast doubt on the value of large AI vendors based in the U.S., including Nvidia. Microsoft, Meta Platforms, Oracle, Broadcom, and other tech giants also saw significant drops as investors reassessed AI valuations.

However, despite the initial success, the startup is now faced with mounting pressure from government agencies around the world, allegations of Chinese state-linked social media accounts creating hype around its initial success, and calls for banning it from Apple and Google App Stores.

So, where does DeepSeek, the startup that threatened to disrupt the way we think about AI models and how they are developed, stand today? Let us take a look.

Accusations that the initial hype was manufactured

When DeepSeek launched its R1 LLM in January 2025, the startup made the model’s code freely available for anyone to use, modify, and distribute. This, when combined with the idea that LLMs, which are the foundation of artificial intelligence, can be developed at a fraction of the cost of what had been spent by companies like OpenAI, Google, and Meta, led to a flurry of activity in the stock market. The result: global investors dumped U.S. tech stocks, wiping $593 billion off chipmaker Nvidia's market value in a record one-day loss for any company on Wall Street.

However, as the dust settled around the initial hype, online analysis firm Graphika came out with a report claiming that Chinese state-linked social media accounts amplified narratives celebrating the launch of the startup's AI models. According to their analysis, the accounts involved in the effort, including those of Chinese diplomats, embassies, and state media, amplified media coverage of the launch and promoted the idea that DeepSeek challenged U.S. dominance in the AI sector.

Graphika further shared that it found a small spike in discussion about DeepSeek's advancements on OpenAI's ChatGPT on X immediately after DeepSeek released its models on Jan. 20, followed by a much larger uptick that continued to build over the weekend.

This report raises questions around the role of the Chinese government in the success of the startup, threatening not just its acceptance by other countries but also raising serious concerns around user privacy.

Bans and investigations: the global response to DeepSeek

Following the launch of the R1 model, the U.S. Commerce Department began looking into whether DeepSeek used U.S. chips that are not allowed to be shipped to China under existing trade restrictions meant to stop Nvidia’s sophisticated AI chips from reaching China.

This, though, was just the beginning.

By June 2025, reports alleged that DeepSeek was aiding China’s military and intelligence operations. Further, U.S. officials claimed that the Chinese tech startup sought to use Southeast Asian shell companies to access high-end semiconductors that cannot be shipped to China under U.S. rules.

The U.S. government also raised concerns around DeepSeek’s privacy policy, claiming it was sharing user information and statistics with Beijing's surveillance apparatus. The claims stemmed from the Chinese law requiring companies operating in the country to provide data to the government when requested.

However, even before the allegations surfaced, governments around the world had increased scrutiny of DeepSeek.

In early February, Australia banned DeepSeek from all government devices over concerns that it posed security risks. South Korea’s data protection authority suspended new downloads of the DeepSeek app after the startup acknowledged failing to take into account some of the agency's rules on protecting personal data.

Similar limitations are also being pursued by other countries. Germany has asked Apple and Google to remove DeepSeek from their stores due to concerns about data safety. India has asked employees of its finance ministry to avoid using AI tools, including ChatGPT and DeepSeek, citing risks posed to the confidentiality of government data. Following India’s lead, Taiwan has also banned government departments from using DeepSeek’s services. The Netherlands, meanwhile, said it would investigate DeepSeek’s data collection practices and urged users to exercise caution when using the company’s software. DeepSeek also faced suspension in South Korea after the country’s data protection authority said that the Chinese artificial intelligence startup transferred user information and prompts without permission.

Sanctions stalled progress

As countries around the world laid down restrictions on the use of DeepSeek and its model, fresh export curbs by the U.S. prevented Nvidia from selling in the Chinese market its H20 chips - the only AI processors it could legally export to the country at the time. And the curbs seem to be having the desired effect.

Fears that the fast adoption of its upcoming R2 could be made difficult due to a shortage of Nvidia server chips in China, and the possibility that a potential surge in demand for R2 would overwhelm Chinese cloud providers, who need advanced Nvidia chips to run AI models, have played a part in delaying its release.

Further, it has been reported that unsatisfactory performance of the reasoning model has also played a role in stalling its release.

What’s next for the Chinese startup

The year 2025 started on a high note for the Chinese AI industry. DeepSeek’s meteoric rise showcased China’s growing prowess in artificial intelligence, but the intense backlash has laid bare the challenges of scaling globally amid a fractured geopolitical landscape.

The startup’s R1 model shook the U.S. tech industry, prompting President Trump to call it a “wakeup call,” and Silicon Valley venture capitalist Marc Andreessen to declare it a “Sputnik moment” in the emerging global AI race. But, with its R2 model stalled, regulatory heat mounting, and critical chip supplies disrupted, DeepSeek’s future now hinges on whether it can build trust across borders while navigating tight domestic and international constraints.

If it succeeds, it could still reshape the AI landscape, but it will need more than just speed; it will need transparency, resilience, and a strategy that goes beyond outperforming models to earning global legitimacy.

Quick Hits

• Cloudflare launches “Bot Paywall” marketplace — Publishers can meter, price, and audit AI scrapers that lift site content, turning bot traffic into a fresh revenue stream.

• Microsoft’s new diagnostic AI outperforms doctors — In early tests, the model beat physicians on complex case accuracy, hinting at faster, lower-cost triage for clinics worldwide.

• Apple gives Siri a ChatGPT upgrade — Upcoming iOS update pipes OpenAI smarts into the voice assistant, promising richer answers and on-device AI summaries.

Prompt of the Day

Little Trivia

Rate this edition