KRX:005380 30 June 2025 - 22 January 2026

Major advances in generative AI continue to emerge. OpenAI has delayed the release of its first open-weight model – initially expected this summer – to allow for more extensive safety checks. CEO Sam Altman said they “need time to run additional safety tests and review high-risk areas,” noting that “once weights are out, they can’t be pulled back” Techcrunch Techcrunch. This pause comes as anticipation builds for OpenAI’s next flagship GPT-5 model, expected later in the year Techcrunch. Meanwhile, the open-source arena is heating up: China’s startup Moonshot AI unveiled Kimi K2, a one-trillion-parameter model that reportedly outperforms OpenAI’s GPT-4.1 on several coding benchmarks Techcrunch – a sign of rising global competition in AI research. Tech giants are pushing generative tech into new domains. At Google I/O 2025, Google announced major updates to its image and video generation tools. It rolled out Imagen 4, a more advanced text-to-image model that handles text better and offers flexible aspect ratios, and Veo 3, a next-gen AI video generator capable of producing video with sound Theverge Theverge. Google also launched “Flow,” an AI filmmaking app that uses Imagen, Veo, and its Gemini AI to create short video clips from text or image prompts