2/6/2025
SoftBank and OpenAI, EU Invests, Fundraising Deal…
Google Removes AI Weapons Ban from Its Principles
Google has revised its AI principles, removing a commitment to avoid using AI for weapons and surveillance. This change aligns with its ambitions in the global AI race but raises ethical concerns.
Source: The AI Track
OpenAI Launches Deep Research for Pro Users
OpenAI introduced Deep Research, a new ChatGPT feature that autonomously conducts complex web research, producing detailed reports in minutes. This tool is designed for professionals and everyday users, leveraging advanced reasoning and data analysis.
Source: The AI Track
SoftBank and OpenAI Form SB OpenAI Japan
SoftBank and OpenAI have launched a joint venture, SB OpenAI Japan, with a $3 billion annual investment. The venture aims to integrate AI technologies into SoftBank’s subsidiaries and provide advanced AI services to major Japanese companies.
Source: The AI Track
EU Invests $56M in OpenEuroLLM
The European Union has announced a $56 million investment to develop OpenEuroLLM, an open-source large language model supporting all 30 European languages. The project leverages EU supercomputers and aims to create models fine-tuned for sectors like healthcare and banking.
Source: Radical Data Science
Anthropic Develops Constitutional Classifiers to Prevent Jailbreaks
Anthropic has introduced a new method to defend AI models against universal jailbreaks. The system uses AI-generated training data in multiple languages and writing styles, achieving a 95.6% success rate in blocking advanced jailbreak attempts.
Source: Radical Data Science
Amazon to Unveil AI-Powered Alexa
Amazon is preparing to launch a smarter, AI-powered version of Alexa, capable of remembering user preferences and handling multiple requests in a single conversation. The new Alexa will rely on AI software from Anthropic and may be offered as a paid service in the future.
Source: iPhone in Canada
DeepSeek Faces Bans in Italy and Australia
Italy’s regulator has blocked the Chinese AI app DeepSeek over data protection concerns, while Australia has banned it on government devices citing security risks. These moves highlight growing scrutiny of AI technologies in global markets.
Source: Gadgets 360
Google’s Gemini 2.0 Expands for Developers
Google has updated its Gemini 2.0 family of models, making Gemini 2.0 Flash generally available to developers. The company also introduced Gemini 2.0 Pro Experimental, its most capable model yet, and Flash-Lite, a low-cost offering.
Source: Radical Data Science
9. OpenAI Nears $40 Billion Fundraising Deal
OpenAI is set to raise 40billioninfunding,pushingitsvaluationto40billioninfunding,pushingitsvaluationto340 billion. This move underscores the ongoing AI investment boom and the increasing stakes in artificial intelligence innovation.
Source: The AI Track
10. DeepSeek’s R1 Model Gains Attention
DeepSeek’s R1 model, which combines multi-stage training and reinforcement learning, has achieved performance comparable to OpenAI’s GPT-4o. The model is seen as a strong competitor in the AI space, particularly in reasoning tasks.
Source: Yicai