Yazılar

Singapore Charges Three Men in Connection with Fraud Against Dell and Super Micro

Singaporean authorities have added additional charges against three men involved in a larger investigation into server fraud, which may involve AI chips. The charges are related to fraudulent activities committed against tech companies Dell and Super Micro. According to court documents, the men allegedly misrepresented the destinations of the servers they purchased, falsely claiming that the servers would not be transferred to unauthorized third parties.

The police charge sheets indicated that the defendants made false representations regarding the ultimate consignee of the items, suggesting that they would not be diverted. Investigations revealed that the servers may contain Nvidia chips, though authorities have not confirmed if these chips are subject to U.S. export controls.

This case is part of an ongoing investigation in Singapore, which involves 22 individuals and companies accused of making false representations. Additionally, the United States is investigating whether DeepSeek, a Chinese company behind a highly publicized AI model, has been using U.S. chips prohibited from being sold to China.

The investigation also covers Malaysia, where the servers allegedly ended up, and authorities there are exploring whether any local laws were violated. Singapore has so far charged two suspects, Aaron Woon, 41, and Alan Wei, 49, with conspiring to defraud Super Micro in 2024. Their charges have been amended to include fraud against Dell. A third suspect, Chinese national Li Ming, 51, was also charged with defrauding Super Micro in 2023.

Authorities seized 42 devices, including phones, computers, and laptops, which are being analyzed by forensic experts. The investigation is still in its early stages, with prosecutors seeking bank statements to trace the movement of funds and planning to work with international law enforcement.

Malaysia Investigates Server Shipments Tied to Singapore Fraud Case

Malaysia has launched an investigation into whether local laws were violated in the shipment of servers linked to a fraud case in Singapore, as the equipment may have contained advanced artificial intelligence (AI) chips subject to U.S. export restrictions.

The case, which surfaced late last month, led Singaporean authorities to charge three men with fraud. Reports suggest the case involves the transfer of Nvidia AI chips to Chinese AI company DeepSeek. Singapore stated that the servers, supplied by U.S. firms, were shipped to Malaysia and may have contained Nvidia chips. However, it has not clarified whether these components fall under U.S. export controls.

Malaysia’s trade ministry issued a statement confirming that the government is taking “necessary actions” to determine if any Malaysian laws were breached regarding the shipment of U.S.-sanctioned AI chips. It further emphasized ongoing collaboration with both the United States and Singapore to develop effective measures for addressing trade compliance issues involving restricted technology.

Meanwhile, U.S. authorities are investigating whether DeepSeek has been using restricted American-made chips to develop its AI models. The Chinese firm’s advancements in AI have drawn significant global attention, particularly after its model’s performance made headlines in January.

Honor Integrates DeepSeek-R1 AI Model with Yoyo Assistant in China

Honor, the prominent Chinese smartphone brand, has officially integrated its Yoyo virtual assistant with the advanced DeepSeek-R1 artificial intelligence (AI) model, making it the first Android manufacturer to incorporate this cutting-edge AI technology into its devices. This integration will enable Yoyo to utilize the DeepSeek-R1 model’s powerful reasoning capabilities to handle complex tasks and deliver a more sophisticated user experience. This marks a significant milestone for Honor as they continue to push the boundaries of AI and virtual assistant functionalities. The DeepSeek-R1 model, which has also been adopted by platforms like Perplexity and Microsoft’s Azure AI Foundry, enhances the ability of Yoyo to perform beyond the usual capabilities of typical AI assistants.

The integration was announced via Honor’s official Weibo account, where the company highlighted the new features that Yoyo users can now access. The update is available to users running MagicOS 8.0 or newer versions, with Yoyo assistant 80.0.1.503 or later. Users in China will be able to access DeepSeek-R1 by visiting the Yoyo smart store, where they can initiate interactions with the model. The welcome screen for DeepSeek-R1 introduces the AI model as an “intelligent agent” based on the open-source DeepSeek-R1, designed to offer users a deeply immersive experience. The model aims to help users explore knowledge, solve problems, and ignite creativity by providing insights and deeper understanding in various contexts.

However, the integration is still in its beta phase, which means that there could be some bugs or performance issues during the initial rollout. Users may experience minor glitches as the system is refined and optimized. Despite this, Honor is optimistic that the DeepSeek-R1 will significantly enhance the Yoyo assistant’s natural language processing, logical reasoning, and overall ability to comprehend and respond to complex user commands. This advancement is expected to push Yoyo to the forefront of virtual assistant technology in China.

The company also revealed that the Yoyo assistant boasts an impressive user base, with over 1.3 billion monthly active users. This large and growing user base presents a significant opportunity for Honor to refine its AI capabilities and expand its reach. As the integration of DeepSeek-R1 becomes fully operational, users can look forward to an even more intelligent and responsive virtual assistant that promises to elevate their digital experiences.