OpenAI unveils first open-weight language models since 2019

OpenAI has released its first open-weight large language models since 2019, marking a pivotal move in the Artificial Intelligence landscape.

OpenAI has launched its first open-weight large language models in over five years, introducing the gpt-oss series in two different sizes. Resembling the company´s o3-mini and o4-mini models in benchmark performance, these new models break from OpenAI´s usual closed approach by being freely downloadable, modifiable, and operable on local hardware, including laptops. The release arrives after multiple delays that fueled community frustration and solidified OpenAI´s ´ClosedAI´ reputation among critics.

This open release positions OpenAI to reclaim influence in a shifting landscape dominated by Meta´s Llama models and rapidly ascending Chinese open models such as DeepSeek, Kimi K2, and Alibaba’s Qwen series. While Meta appears to be moving toward more closed releases and Chinese models increasingly gain market share, OpenAI aims to fill what it sees as a critical gap for organizations seeking flexible or on-premise language model solutions. The gpt-oss models cater to a range of users, including enterprises prioritizing customization or data security, as well as researchers who require access to model internals for scientific inquiry.

Crucially, OpenAI has chosen the permissive Apache 2.0 license, allowing unrestricted commercial use and aligning its release more closely with Chinese models than the more restrictive terms applied by Meta. Experts like Nathan Lambert and Peter Henderson highlight the potential for enhanced research collaboration and domestic technological leadership, emphasizing that open models serve as a key tool in both academic research and broader technological strategy. The timing also dovetails with political priorities: the Trump administration´s AI Action Plan emphasizes open models, and OpenAI´s alignment could secure valuable governmental support as it scales up its operations.

Geopolitics underscores the move. Observers note the content limitations and potential risks associated with some Chinese models, and influential researchers advocate for sustained American investment in transparent, open Artificial Intelligence. OpenAI´s initiative is thus as much about reasserting leadership in the global AI research ecosystem as it is about meeting customer demand. The company, in its announcement, stressed the broader democratic and technological stakes, positioning American-created open-weight models as critical infrastructure for the global Artificial Intelligence community.

82

Impact Score

UK and EU Artificial Intelligence regulatory outlook for May 2026

The UK is moving ahead with targeted Artificial Intelligence measures in policing, online safety, cyber security and copyright policy, while the EU is refining how the EU Artificial Intelligence Act will apply in practice. Consultations, new offences and implementation deadlines are shaping the next phase of compliance on both sides.

Germany sets out national implementation of the Artificial Intelligence Act

Germany has published a draft law to implement the European Artificial Intelligence Act through new supervisory structures, clearer institutional responsibilities, and measures designed to support innovation. The proposal puts the Federal Network Agency at the center of enforcement while preserving sector-specific oversight in sensitive fields.

ECB warns banks about new Artificial Intelligence security risks

The European Central Bank has called major banks to an emergency meeting over cybersecurity risks tied to advanced Artificial Intelligence models. Regulators want banks to speed up security updates as newer tools make it easier to find and exploit vulnerabilities.

Contact Us

Got questions? Use the form to contact us.

Contact Form

Clicking next sends a verification code to your email. After verifying, you can enter your message.