Episode Summary: In this episode of the Vertica Newshound, we explore the latest groundbreaking strides in artificial intelligence, with a focus on the newly unveiled 'Deep Research' AI agent by OpenAI. We also examine significant AI governance developments from the European Union's regulatory stance to security vulnerabilities, alongside strategic AI initiatives in Japan.Main Story: OpenAI Unveils 'Deep Research' AI Agent for Enhanced Online ResearchOpenAI has announced 'Deep Research,' their latest AI agent designed to revolutionize online research. This tool mimics a seasoned analyst, condensing hours of investigation into minutes by autonomously extracting and compiling data from diverse sources. Despite its potential, 'Deep Research' faces challenges like "hallucination" issues and high computational demands, limiting initial availability to premium users. The launch comes amid fierce competition, particularly from China's DeepSeek, as both entities seek dominance in AI innovation.Sources:- [Bloomberg](https://www.bloomberg.com/news/articles/2025-02-03/openai-releases-ai-agent-aimed-at-speeding-up-online-research)- [The Verge](https://www.theverge.com/news/604902/chagpt-deep-research-ai-agent)- [The New York Times](https://www.nytimes.com/2025/02/02/technology/openai-deep-research-tool.html)- [Axios](https://www.axios.com/2025/02/02/chatgpt-deep-research-openai-agents)- [Times of India](https://timesofindia.indiatimes.com/world/rest-of-world/openai-unveils-deep-research-tool-as-chinas-deepseek-heats-up-ai-race/articleshow/117872763.cms)- [The Guardian](https://www.theguardian.com/technology/2025/feb/03/deepseek-chinese-ai-company-competitors-chatgpt)- [CNN](https://www.cnn.com/2025/02/02/uk/britain-criminalizes-ai-child-sex-abuse-intl-hnk/index.html)Secondary Stories:1. AI Systems with ‘Unacceptable Risk’ Are Now Banned in the EUThe European Union enforces its AI Act, banning systems deemed to pose "unacceptable risk" to ensure ethical AI practices. This regulatory move requires businesses in the EU to align their AI deployment strategies with strict compliance standards.[Read more at TechCrunch](https://www.techcrunch.com/2025/02/02/ai-systems-with-unacceptable-risk-are-now-banned-in-the-eu/)2. DeepSeek's AI Model Faces 100% Attack Success Rate in Security TestsResearchers reveal that DeepSeek's chatbot security is highly vulnerable, achieving a 100% success rate in simulated attacks. This underscores the critical need for robust security measures in AI applications to prevent exploitation.[Read more at Interesting Engineering](https://interestingengineering.com/innovation/deepseek-fail-not-block-harmful-prompt)3. SoftBank's Son and OpenAI's Altman to Unveil AI Initiative in JapanMasayoshi Son and Sam Altman announce a new AI venture in Japan, focusing on enhancing AI infrastructure. This initiative reflects a global trend toward investing in AI capabilities, likely influencing corporate AI strategies, especially in e-commerce.[Read more at Bloomberg](https://www.bloomberg.com/news/articles/2025-02-03/softbank-s-son-openai-s-altman-to-unveil-japan-ai-thrust)4. UK to Outlaw AI Tools for Child Abuse in Groundbreaking LegislationThe UK introduces pioneering laws banning AI tools used in creating child sexual abuse content, setting a precedent for international legal frameworks. This move highlights the importance of ethical AI usage and robust regulations to prevent misuse.[Read more at CNN](https://www.cnn.com/2025/02/02/uk/britain-criminalizes-ai-child-sex-abuse-intl-hnk/index.html)Stay informed and leverage these insights to navigate the evolving AI landscape.