#OpenAI

OpenAI has launched three new advanced AI models: GPT-5.6 Sol, Terra, and Luna. Access is initially restricted to trusted partners following a U.S. government request. OpenAI aims for broader public release soon, emphasizing collaboration with the government on pre-release assessments. GPT-5.6 Sol is their most sophisticated model, with notable advancements in coding, biology, and cybersecurity. This move highlights the evolving balance between AI innovation and governmental oversight.

AI spending is increasing as companies race to integrate AI. This has led to surging costs, with some reporting astronomical bills. In response, some companies are shifting to more cost-effective alternatives, while others are implementing spending caps. OpenAI and Anthropic, the leading AI model developers, are reportedly preparing for IPOs amidst growing demand for ROI and competition from giants like Microsoft, Amazon, and Google, who are developing their own more affordable AI models.

OpenAI is developing its own custom “Jalapeño” chip with Broadcom to address the immense costs of scaling AI models like ChatGPT. This move aims to improve efficiency and reduce reliance on third-party hardware, mirroring Apple’s vertical integration strategy. The chip is optimized for LLM inference, minimizing data movement. OpenAI’s accelerated chip design process leverages its own AI models, with initial deployment planned for late 2026.

Broadcom’s stock may find a bottom following the unveiling of its custom AI accelerator chip, “Jalapeno,” co-designed with OpenAI. This collaboration aims to boost AI speed, reliability, and accessibility while reducing costs through enhanced performance per watt. Despite recent stock pressure and competition, this partnership offers a potential catalyst for Broadcom, with initial Jalapeno deployments expected later this year and significant ramp-up in 2027. The chip is seen as complementary to existing GPU infrastructure, addressing the industry’s “insatiable” demand for computing power.

OpenAI and Broadcom have launched “Jalapeño,” a custom AI chip designed in nine months to power OpenAI’s inference operations. This strategic move into the AI silicon market allows OpenAI to build its own infrastructure for greater efficiency and accessibility. The chip, an ASIC, is optimized for specific AI tasks and aims to make advanced AI faster and more reliable. Initial deployment is targeted for late 2026, with full-scale operations expected in 2028.

Samsung Electronics is expanding its use of OpenAI’s ChatGPT Enterprise and Codex to all employees, including the global Device eXperience division. This move aims to boost innovation and efficiency across various functions like software development, marketing, and product ideation, enhancing tasks from information retrieval to code debugging. The initiative signifies a secure integration of generative AI, contrasting with previous data security concerns, and aligns with Samsung’s role as a strategic memory partner for OpenAI’s infrastructure.

L’Oréal partners with OpenAI to integrate Maybelline’s virtual makeup try-on into ChatGPT, enhancing consumer engagement and product discovery. This collaboration spans AI-native advertising pilots and advanced internal research, utilizing models like GPT-Rosalind for scientific discovery. The initiative aims to accelerate product formulation and empower employees with generative AI tools, positioning L’Oréal at the forefront of beauty tech innovation.

Microsoft has become China’s primary provider of advanced OpenAI AI models, leveraging its Azure cloud infrastructure. Despite OpenAI and Anthropic avoiding direct engagement due to IP concerns, Microsoft supplies models to Chinese giants like ByteDance, Ant Group, Meituan, and Tencent. This strategy generates significant revenue and positions Microsoft as a unique intermediary, navigating geopolitical risks and offering both American and Chinese AI models.

Noam Shazeer, a key figure behind Google’s Gemini AI models, has joined OpenAI. This move highlights the fierce competition for AI talent. Shazeer previously co-founded Character.AI after leaving Google. His departure, occurring shortly after Google’s AI product launches, underscores the ongoing talent migration in the AI industry.

Meta has re-entered the AI arena with a significant investment and the integration of Scale AI engineers, led by Alexandr Wang. Their new proprietary model, Muse Spark, signals a shift from open-source to internal applications. While Meta faces challenges in competing with rivals and regaining developer trust after a misstep with Llama, CEO Mark Zuckerberg must now demonstrate tangible financial success from AI-driven tools beyond ad optimization. Investors are watching closely for monetization of new AI products.