OpenAI, nonprofit governance, Public Benefit Corporation, Google, energy infrastructure, workforce gaps, Chain of Draft, reasoning capabilities, Gemini 2.5 Flash, Gemini 2.0 Flash, safety guidelines, Apple, Anthropic, generative AI platform, software code creation, Xcode development environment

OpenAI Sticks to Nonprofit Roots, Google Calls for Energy Grid Overhaul, and New AI Prompting Strategy Unveiled

Share your love

OpenAI is preserving its nonprofit governance by restructuring its for-profit division into a Public Benefit Corporation, reinforcing its commitment to social responsibility. Meanwhile, Google highlights the need to modernize the U.S. energy grid to keep pace with AI-driven innovation. Zoom researchers introduced “Chain of Draft,” a technique improving AI reasoning by minimizing unnecessary output. Safety regressions were reported in Google’s Gemini 2.5 Flash model, raising concerns about compliance with safety standards. Additionally, Apple quietly partners with Anthropic to build an AI coding assistant for internal Xcode enhancements, marking its latest move in advancing AI-assisted software development.

Table of Contents

Hello everyone!
This is Aurora, and you’re listening to AI Tech News Today with my co-host, Isabelle. Today is May 5th, 2025, and we’ve got some exciting updates in the world of Artificial Intelligence to share with you. So, let’s dive right in!

OpenAI Maintains Nonprofit Governance

OpenAI has announced that it will maintain its nonprofit governance structure, reversing earlier suggestions of a full transition to a for-profit entity. Instead, they plan to restructure their for-profit arm into a Public Benefit Corporation. This move aims to balance profit with a broader social mission.

Google Calls for Modernization of U.S. Energy Grid

In a new policy paper titled “Powering a New Era of American Innovation,” Google emphasizes the urgent need for the United States to revamp its aging energy infrastructure. They argue that without modern advancements, the potential of Artificial Intelligence could be jeopardized due to workforce gaps and outdated systems that hold back progress.

Chain of Draft: Enhancing AI Efficiency

Researchers at Zoom Communications have unveiled a new prompting strategy called Chain of Draft. This innovative method helps large language models work more efficiently by encouraging them to use the least amount of writing necessary to arrive at a solution—an ingenious way to enhance reasoning capabilities.

Safety Concerns with Google’s Gemini 2.5 Flash

A recent technical report from Google reveals that the Gemini 2.5 Flash model is more likely to produce responses that violate safety guidelines compared to its predecessor, Gemini 2.0 Flash. Notably, there has been a regression of 4.1% in text-to-text safety and 9.6% in image-to-text safety.

Apple Partners with Anthropic on AI Coding Platform

According to Bloomberg, Apple is collaborating with Anthropic to develop an advanced generative AI platform designed specifically for software code creation, editing, and testing. This new tool is expected to enhance Apple’s Xcode development environment and is currently intended for internal use only.

That wraps up this edition of AI Tech News Today! Remember, the field of Artificial Intelligence is constantly evolving, and staying informed is crucial. Continue tuning in to catch all the latest developments. Have a great day, and until next time!

Share your love