OpenAI's new Ghibli-style filter on its ChatGPT platform has sparked a viral trend, with users creating and sharing images in the style of Studio Ghibli classics like "My Neighbor Totoro" and "Princess Mononoke". The tool, powered by GPT-4o, allows for sophisticated results through concise requests. However, the sudden popularity has raised concerns about copyright infringement, prompting lawsuits against OpenAI. The company's CEO, Sam Altman, acknowledged the unexpected success and announced that the tool would remain limited to paid users for now due to "GPU melting" issues.
https://www.news18.com/tech/chatgpts-ghibli-craze-online-dogged-by-copyright-lawsuits-9277769.htmlThe Open Web Application Security Project (OWASP) has expanded its focus to include additional resources for CISOs and compliance officers related to Large Language Model (LLM) application security, with a broader scope that now includes governance, risk management, and compliance. The OWASP Gen AI Security Project, led by experts from over 18 countries, 130 companies, and 8,000 community members, has grown to address the security challenges in autonomous systems, including agentic AI applications. A new guide provides developers and security professionals with a structured taxonomy and mitigation strategies for emerging threats in agentic AI applications. The project's expansion reflects its maturity and strategic alignment with OWASP's core mission, and it has been promoted to a flagship project by the Global Board of Directors.
https://www.prnewswire.com/news-releases/project-owasp-promotes-genai-security-project-to-flagship-status-302412689.htmlSam Altman's profile picture was changed to an AI-generated Ghibli-style version of himself after receiving hundreds of messages from users who had created a "twink" version of him. OpenAI's latest update, powered by GPT-4o, allows users to generate images directly within the chatbot, but free-tier users face limitations due to high demand. The new feature has improved image accuracy and text rendering, although it creates images piece by piece unlike DALL-E, which generates them all at once.
https://www.ndtv.com/world-news/openai-chatgpts-why-social-media-is-suddenly-flooded-with-studio-ghibli-style-art-after-gpt-4o-launch-8020875#pfrom=home-ndtv_featured_featuredImgSANS is creating a critical moment in the industry by releasing Critical AI Security Guidelines, hosting an AI Cybersecurity Hackathon, and offering collaborative education at the AI Summit. Professionals can unite, innovate, and build the future of secure AI through these initiatives, with pre-registration available for the guidelines and registration open for the AI Summit 2025.
https://www.manilatimes.net/2025/03/27/tmt-newswire/globenewswire/building-a-safer-ai-future-sans-leads-coordinated-effort-to-secure-artificial-intelligence/2081465The US Securities and Exchange Commission (SEC) has warned about the growing sophistication of financial fraud perpetuated by artificial intelligence (AI). Agentic AI tools can work autonomously to help scammers defraud financial companies, making it difficult to trace the crimes. Experts like Katherine Forrest from Paul Weiss law firm noted that these AI models have "baton hand-offs" between them, highlighting the complexity of tracing such fraud.
https://news.bloomberglaw.com/banking-law/agentic-ai-presents-new-frontier-for-fraud-against-financial-companiesSK Hynix, South Korea's second-largest memory chipmaker, is preparing for potential US tariffs on semiconductor imports, with some customers bringing forward orders to stock up before the tariffs take effect. The company expects "explosive growth" in high bandwidth memory (HBM) chips demand this year, driven by investments in data centers, and its CEO Kwak Noh-Jung sees DeepSeek's emergence as a positive development for SK Hynix, potentially boosting medium-to-long-term demand for AI memory chips.
https://channelnewsasia.com/business/sk-hynix-says-some-customers-brought-forward-orders-ahead-us-tariffs-5026881Socure, a leading provider of artificial intelligence for digital identity verification, has achieved "Moderate" authorization from the Federal Risk and Authorization Management Program (FedRAMP) for its Socure For Government platform. This milestone enables federal agencies to adopt advanced digital identity verification capabilities to combat identity fraud and improve government efficiency. With FedRAMP approval, Socure's platform can analyze and correlate every facet of an individual's digital identity in real-time, detecting and preventing fraud before it occurs. The platform offers auto-approval for 99% of mainstream populations and 94% for hard-to-identify populations, capturing 99% of third-party identity fraud attacks.
https://www.prnewswire.com/news-releases/socure-earns-moderate-fedramp-authorization-302413185.htmlOpenAI is close to finalizing a $40 billion funding round led by SoftBank Group Corp., with investors including Magnetar Capital, Coatue Management, Founders Fund and Altimeter Capital Management in talks to participate, according to people familiar with the matter. The funding round could be the largest of all time, valuing OpenAI at $300 billion.
https://news.bloomberglaw.com/artificial-intelligence/openai-close-to-finalizing-its-40-billion-softbank-led-fundingOpenAI is close to finalizing a $40 billion funding round led by SoftBank Group Corp., with investors including Magnetar Capital, Coatue Management, Founders Fund and Altimeter Capital Management in talks to participate, according to people familiar with the matter.
https://news.bloomberglaw.com/mergers-and-acquisitions/openai-close-to-finalizing-its-40-billion-softbank-led-funding