Connect with us

Hi, what are you looking for?

Top Stories

Google Launches Open-Source Gemma 4 with Apache 2.0 License After Developer Exodus

Google shifts to open-source with the launch of Gemma 4 under the Apache 2.0 license, enabling unrestricted commercial use amid rising competition.

Google has pivoted towards an open-source strategy by launching its lightweight AI model, Gemma 4, under the Apache 2.0 license, which allows unrestricted commercial use. This move, announced recently, is seen as a response to losing developers to competitors like DeepSeek and Qwen. Analysts suggest this shift indicates a course correction as U.S. tech giants have largely maintained closed ecosystems, while European AI startups have leveraged these Chinese models to create enterprise agents and chatbots.

Gemma 4, constructed on similar technology as the previously launched Gemini 3, is designed to operate offline on devices such as smartphones and laptops. By opening up its model, Google aims to capture a potentially lucrative market for on-device AI agents across billions of Android devices.

In addition to Google’s strategic shift, the Korea Exchange has raised the threshold for initial public offerings (IPOs), mandating treasury share cancellations before preliminary listing reviews. This new requirement effectively halts the traditional practice of funding through the sale of old treasury shares. Analysts note that with this change, coupled with increasing tax implications from rising controlling-shareholder stakes, startups will likely need to rethink their listing strategies significantly.

On the startup front, a new development is taking shape in Northwestern Seoul. Plans are in place for a bio-research complex and startup spaces on approximately 50,000 pyeong (about 165,000 square meters) of land made available by the undergrounding of the Gyeongui Line. This initiative seeks to create a “mini Silicon Valley,” bolstered by a robust student population of around 12,000 international students from institutions like Yonsei University, Ewha Womans University, and Myongji University.

A variety of initiatives are emerging to support startups. Samsung SDS has introduced a new consulting program called S-VIP, aimed at guiding manufacturing companies through their AI transformation (AX) processes. This program unfolds in three stages: establishing AI objectives, identifying use cases, and building governance frameworks. In a pilot project with a major rental company, technicians reduced search time by 43.9% while improving accuracy by 12.45%, indicating strong potential for further investment in AI solutions.

The Ministry of Planning and Budget is also stirring interest among startups by instituting a new performance bonus scheme. Public-sector officials can now receive bonuses of up to 10 million won, intended to incentivize a more performance-driven culture. This initiative is expected to foster new collaboration opportunities between startups and government projects as various ministries explore similar incentive programs.

Lastly, the vision of transforming the Sinchon area into a hub for startups continues to gain momentum. Plans to develop an international young entrepreneurs’ city have attracted interest from several companies, with expectations that building height restrictions along key roads may soon be relaxed to facilitate development. A feasibility study for a mixed-use project in front of Yonsei University’s main gate is also underway, suggesting potential synergies with the developments associated with the Gyeongui Line.

In summary, the landscape for AI and startups in South Korea is evolving rapidly. As Google repositions itself in the open-source arena, the regulatory environment shifts, and new urban development plans emerge, the implications for innovation and entrepreneurship in the region remain significant.

See also
Staff
Written By

The AiPressa Staff team brings you comprehensive coverage of the artificial intelligence industry, including breaking news, research developments, business trends, and policy updates. Our mission is to keep you informed about the rapidly evolving world of AI technology.

You May Also Like

AI Regulation

China's OpenClaw initiative introduces a comprehensive AI governance framework, aligning ethical regulations with national interests to foster responsible innovation.

Top Stories

DeepSeek is set to launch its V4 AI model, enhanced by Huawei's latest chips, with major orders from Alibaba, ByteDance, and Tencent fueling anticipation.

AI Cybersecurity

OpenAI acquires Promptfoo for enhanced AI security capabilities, integrating cutting-edge tools used by 25% of Fortune 500 companies into its Frontier platform.

AI Marketing

Criteo launches Criteo GO, a generative AI tool enabling SMBs to create ad campaigns in five clicks, achieving over 20% higher ROI than traditional...

AI Technology

Google unveils TurboQuant at ICLR, promising significant AI inference performance boosts on existing hardware without costly upgrades or architectural changes

Top Stories

DeepSeek forecasts Nvidia's stock will surge 50% to $265 by 2026, driven by new technology and strong institutional confidence amid market challenges.

AI Generative

Google launches Gemma 4, an open-source AI suite with 26B and 31B models for local deployment, enhancing privacy and multimodal reasoning capabilities.

AI Research

Google's TurboQuant breakthrough slashes memory usage by 600% and enhances attention computation by 800%, transforming AI efficiency and market dynamics.

© 2025 AIPressa · Part of Buzzora Media · All rights reserved. This website provides general news and educational content for informational purposes only. While we strive for accuracy, we do not guarantee the completeness or reliability of the information presented. The content should not be considered professional advice of any kind. Readers are encouraged to verify facts and consult appropriate experts when needed. We are not responsible for any loss or inconvenience resulting from the use of information on this site. Some images used on this website are generated with artificial intelligence and are illustrative in nature. They may not accurately represent the products, people, or events described in the articles.