Yazılar

DeepSeek Limits Chip Access

/in /tarafından

Chinese AI developer DeepSeek has reportedly withheld early access to its upcoming model from major U.S. chipmakers, marking a departure from common industry practices.

Typically, AI labs collaborate with hardware providers before releasing new systems to ensure optimal performance across widely used processors. In this case, domestic suppliers were granted priority access to prepare for deployment.

The decision comes amid intensifying competition in the global AI sector, where technological alignment between software and hardware plays a critical role in efficiency.

Analysts suggest that advances in development tools may reduce reliance on extended pre-release optimization, potentially minimizing operational impact on external hardware partners.

The move highlights shifting dynamics in the AI ecosystem as regional technology strategies evolve and competition for leadership in advanced computing continues.

DeepSeek Plans February Launch of Coding-Focused AI Model V4, Report Says

/in /tarafından

Chinese artificial intelligence startup DeepSeek is expected to launch its next-generation AI model, V4, in mid-February, with a strong focus on coding capabilities, The Information reported on Friday, citing people familiar with the matter.

According to the report, internal testing by DeepSeek employees indicates that the V4 model could outperform leading rivals such as Anthropic’s Claude and OpenAI’s GPT series in software development and coding-related tasks. The model is also said to show major improvements in handling extremely long and complex coding prompts, a feature that could give developers an edge when working on large-scale or intricate software projects.

Reuters was not able to independently verify the report, and DeepSeek did not immediately respond to a request for comment.

Based in Hangzhou, DeepSeek has rapidly emerged as a prominent player in China’s drive to build a domestic AI ecosystem and reduce reliance on foreign technology. The company attracted international attention after several Silicon Valley executives publicly praised its earlier models, including DeepSeek-V3 and DeepSeek-R1, for their performance and cost efficiency.

In January, Reuters reported that DeepSeek claimed to have developed a low-cost rival to ChatGPT, further boosting its profile in the global AI race. However, the company has also faced scrutiny in some countries over concerns related to data security and privacy practices, highlighting the growing regulatory attention surrounding advanced AI systems.

Italy Closes Probe Into DeepSeek After Commitments to Warn Users of AI “Hallucination” Risks

/in /tarafından

Italy’s antitrust authority has closed an investigation into Chinese artificial intelligence system DeepSeek after the company agreed to binding commitments aimed at improving warnings about the risk of AI-generated false information.

The probe, launched last June by Italy’s antitrust and consumer protection authority AGCM, focused on allegations that DeepSeek failed to adequately inform users that its AI system could generate inaccurate, misleading, or fabricated content — commonly referred to as “hallucinations.”

The decision to end the investigation was announced in the AGCM’s weekly bulletin published on Monday. According to the regulator, the commitments were submitted by Hangzhou DeepSeek Artificial Intelligence and Beijing DeepSeek Artificial Intelligence, which jointly own and operate the DeepSeek platform.

The agreed measures include clearer and more prominent disclosures explaining the risk that, based on user inputs, the AI model may produce outputs containing incorrect or invented information. The AGCM said the new disclosures are designed to be more transparent, intelligible, and immediately visible to users.

“The commitments presented by DeepSeek make disclosures about the risk of hallucinations easier, more transparent, intelligible, and immediate,” the authority said in its bulletin.

The case highlights growing regulatory scrutiny across Europe over how AI systems communicate their limitations to users, particularly as generative AI tools become more widely adopted in consumer-facing applications.