A Chinese AI company, DeepSeek, has unveiled an open-source AI image generator called Janus-Pro-7B, claiming it outperforms OpenAI's DALL-E 3 and Stability AI's Stable Diffusion. This move sent shockwaves through the US stock market. The new model is a significant upgrade from DeepSeek's previous image generator, Janus, launched in late 2024.
https://petapixel.com/2025/02/26/alibaba-makes-ai-video-generator-wan-2-1-free-to-use/Anthropic's new AI model is valued at $61.5 billion after securing a $3.5 billion funding round, with Amazon investing $8 billion and Alphabet investing $2 billion. The company's Claude model can now command computers like humans do, following the release of OpenAI's ChatGPT six months prior.
https://www.hurriyetdailynews.com/amazon-backed-anthropic-releases-its-smartest-ai-model-206221Anthropic's new AI model, Claude, has improved its coding capabilities and can now handle complex tasks more effectively than its predecessor. The company's co-founder Jared Kaplan attributes this improvement to a "hybrid" reasoning model that allows for quick answers or in-depth analysis of queries. This advancement enables Claude to better follow instructions and perform sophisticated analyses. Anthropic is competing with OpenAI, which recently released an AI agent called Operator, and has received significant investments from Amazon ($8 billion) and Alphabet (Google's parent company, $2 billion), as it strives to stand out in the crowded market.
https://www.geo.tv/latest/592366-openai-rival-anthropic-introduces-its-smartest-ai-modelJustine Roberts, founder of Mumsnet, has warned that media companies are slow to recognize the danger posed by artificial intelligence, particularly from tech giants like OpenAI, which makes ChatGPT. Mumsnet became the first British website to take legal action against OpenAI last year, claiming copyright infringement after its content was scraped without permission to train ChatGPT. Roberts believes proposals allowing AI firms to ignore copyright law are "shortsighted" and could destroy creative industries, as models rely on a constant supply of fresh content.
https://www.dailymail.co.uk/news/article-14432389/Mumsnet-founder-Justine-Roberts-says-creators-present-united-against-existential-threat-AI.htmlOpenAI has removed accounts from China and North Korea suspected of using its technology for malicious activities such as surveillance and opinion-influence operations. The company used AI tools to detect these operations, which include generating news articles in Spanish denigrating the US and creating fictitious job profiles to commit fraud. These actions are believed to be ways authoritarian regimes could leverage AI against their own people and the US.
https://www.deccanchronicle.com/technology/openai-removes-users-in-china-north-korea-suspected-of-malicious-activities-1862732Only 9% of Japanese people have used generative AI, despite 61% being aware of it, according to a recent survey by Nomura Research Institute. The technology has sparked concerns about privacy and security, with the government taking action to restrict its use. In contrast, businesses in labor-strapped industries such as retail are more likely to have adopted generative AI, citing benefits but also bottlenecks like a lack of specialists and accurate information. SoftBank Group CEO Masayoshi Son plans to set up an AI venture in Japan with OpenAI, while the government aims to unveil a master plan to promote AI development and use in a "safe and secure" manner.
https://mainichi.jp/english/articles/20250215/p2g/00m/0sc/021000c