AI News — Sunday, July 12, 2026
OpenAI has launched GPT-Live, a new product offering real-time conversational AI capabilities, signaling a significant advancement in interactive AI.
OpenAI is strategically targeting families, aiming to embed ChatGPT more deeply into daily household life with new features and use cases.
The CEO of Hugging Face highlights a growing trend where companies prefer to own and customize their AI models rather than relying solely on rented services.
Listen Labs secured $69 million in funding following a viral hiring campaign, which will be used to expand its AI-powered customer interview platform.
Deutsche Telekom is actively integrating AI across its operations to innovate and enhance its telecommunications infrastructure and services.
SK Hynix completed the largest foreign IPO in US history, raising $26.5 billion, and faces calls to establish new fabrication plants in the United States.
OpenAI is launching initiatives to equip K–12 educators with practical AI skills, fostering AI literacy and integration in classrooms.
An AI summit in New York City brought together educators and industry experts at Google's offices to discuss and define the role of AI in future education.
A developer shares a method for integrating AI into Slack to automate the creation and submission of pull requests, streamlining development workflows.
DrugGen 2 introduces a new disease-aware language model designed to significantly improve the efficiency and targeting of drug discovery processes.
New research proposes a geometric-aware pretraining method to significantly improve the quality and coherence of in-context panoramic image generation.
Jet-Long presents an innovative approach using dynamic bifocal RoPE to efficiently extend the context window of large language models, improving their ability to handle longer inputs.
JD.com introduces Oxygen AIIC V1, an industrial-scale solution leveraging LLM/VLM for comprehensive item understanding, management, and diverse applications.
This paper introduces a novel single-rollout asynchronous optimization technique to enhance the training efficiency and performance of agentic reinforcement learning systems.
An article explores the phenomenon of AI instruction decay and provides strategies for crafting more robust and enduring prompts for AI models.