DeepSeek reports cyberattack as users make a run for the AI startup

DeepSeek, the Chinese AI startup that came out with an AI assistant that could more than match American AI models like ChatGPT, announced a temporary halt to user registrations following, what the startup said, a cyberattack and a sudden rush for registrations.

In fact, the startup saw a rush of user registrations, leading to outages, after the model became the to-rated free-to-use AI assistant available on Apple’s App Store.

DeepSeek said it resolved the issue of users not being able to log in to the website.

While DeepSeek resolved the temporary glitch, the disruption of the AI market with its low-cost AI assistant, caused a $593 billion dent on Nvidia’s market cap, and send the Nasdaq stock index crashing 3.1 per cent.

DeepSeek shot to fame last month after its models, including the V3 and R1 outperformed similar models of US tech giants, including Nvidia, which prompted President Donald Trump raise an alarm over US laxity in new age technology development.

Researchers at DeepSeek-AI have developed an upgraded version of Janus, Janus-Pro, to overcome the limitations of earlier models. Janus-Pro offers an optimised training strategy, an expanded and high-quality dataset, and larger model variants – Janus-Pro-1B and Janus-Pro-7B.