Inside OpenAI's Third Phase: Personal AGI, an AI Researcher, and a Coordinated Push for Global Safety Oversight

The first two phases of OpenAI's history were about proof of concept and then massive scale. The third phase, as described by Altman and Pachocki, is about democratization. The central goal is no longer just to build the most powerful AI but to make advanced AI abundant, affordable, safe, practical, and sufficient for every person and organization on the planet . The strategy represents a significant pivot from a lab racing for raw capability to a platform company preparing for life on the public markets .

The Three Core Goals of OpenAI's New Roadmap

The company's public plan is built on three specific, ambitious objectives :

1. Build an automated AI researcher. OpenAI aims to create an AI system that can conduct significant scientific research on its own. Altman has pointed to a target of March 2028 for AI to handle a "significant" portion of the company's own research and development work, a milestone that could accelerate the entire field of AI safety and capability .
2. Accelerate the economy through scientific progress. The second pillar ties AI-driven breakthroughs directly to broad economic growth, with an emphasis on distributing the gains widely rather than concentrating them .
3. Give every person on Earth a personal AGI. The most sweeping and direct-to-consumer goal is to make a form of artificial general intelligence personally accessible and useful to every individual, reframing AGI not as a centralized corporate or government tool but as a personal resource .

The Safety Stance: "Entirely Automating Everything Is Not the Future We Want"

Woven into the plan is a distinct set of safety and control principles that serve as a manifesto for a human-centric future. The post states directly that "powerful systems must remain safe, aligned with human intent, and subject to human control" . Perhaps more striking was the explicit reassurance that "entirely automating everything is not the future we want," a direct message intended to counter fears of a fully autonomous, human-free economic future . This language positions OpenAI as a steward of technology designed to augment rather than replace human agency.

The Big Ask: A Global Body with the Power to Pause

Beyond the product and mission statements, the most geopolitically significant element of the plan is its call for global governance. OpenAI expressed strong support for an international body—potentially led by the U.S. and including China—with the explicit authority to coordinate action on frontier AI risks . The company said one goal of such an organization should be to "make it possible for the world to take coordinated action, including slowing frontier development when needed, so societal resilience, safety, and alignment can keep pace" .

This call for a global "pause switch" is a remarkable public stance for a company racing toward a potential trillion-dollar valuation. It formally acknowledges a gap between the speed of technical progress and society's ability to manage it safely .

A Coordinated Industry Moment: Anthropic's Parallel Move

The timing of OpenAI's plan is impossible to separate from its industry context. Just one week earlier, Anthropic, a chief rival, had also confidentially filed for an IPO and released its own public statement declaring it would be "good for the world to have the option to slow or temporarily pause frontier AI development to enable societal structures and alignment research to keep up with the advance of the technology" . Anthropic even announced the creation of the Anthropic Institute, a new entity specifically tasked with building the verification systems that a credible global slowdown would require .

The near-simultaneous IPO filings and safety proposals from the two leading frontier labs created a unique moment . Critics have argued that these coordinated calls for regulation can be framed as competitive positioning—a way for established players to raise the barriers for new entrants while appearing responsible as they navigate the scrutiny of public markets . Regardless of intent, the outcome is the same: the future of AI governance and the biggest tech IPOs in history are now negotiating their terms in public, together.

The Road to the Public Markets

While the "third phase" plan lays out a long-term vision, the immediate business development is procedural but momentous. OpenAI's confidential S-1 filing opens the door for what could be one of the largest public market debuts ever, with Reuters reporting the company could be targeting a valuation of up to $1 trillion as soon as September 2026 . However, Altman has cautioned that "we have not decided on timing yet; it may be a while," acknowledging that some strategic moves are easier to execute while still a private entity . Internal reports have also surfaced of leadership debate on the ideal timing, with CFO Sarah Friar raising concerns about revenue trajectory and public-company reporting standards, all while the company navigated missing several internal monthly revenue targets in the lead-up to the filing .