OpenAI's Third Phase: The Strategic Shift from Research to Distribution

On June 8, 2026, OpenAI CEO Sam Altman and CTO Jakub Pachocki published a manifesto outlining the company's third phase: making advanced AI abundant, affordable, and accessible. The core promise: 'Give everyone on Earth a personal AGI.' This is not a product launch; it is a strategic declaration that redefines OpenAI's competitive posture and the entire AI industry's trajectory.

By March 2028, OpenAI expects a 'significant fraction' of its research to be conducted by AI systems. This timeline is aggressive. It implies that within 18 months, OpenAI will have automated the very process of innovation. The strategic consequence is profound: if AI can accelerate AI research, the rate of progress becomes exponential, and the gap between OpenAI and competitors widens dramatically.

For executives, this means the window to build proprietary AI capabilities is closing. The choice is not whether to adopt AI, but whether to align with OpenAI's ecosystem or risk being left behind. The manifesto's emphasis on 'broad distribution of power' is a direct appeal to regulators and the public, positioning OpenAI as the benevolent steward of AGI. However, the reality is that OpenAI controls the most advanced models, the distribution channels (via partnerships like Dell Technologies for Codex), and the research agenda. This is a classic platform play: standardize, subsidize, and capture the ecosystem.

The Automated AI Researcher: A Double-Edged Sword

OpenAI's first goal is to 'build an automated AI researcher.' This is not merely a tool; it is a force multiplier. By 2028, AI could be generating hypotheses, running experiments, and writing papers. The strategic implication is that OpenAI's research velocity will outpace any human-only lab. Competitors like Google DeepMind, Anthropic, and xAI will face a stark choice: partner with OpenAI, invest heavily in their own automated research, or fall behind.

But there is a hidden risk: alignment. If AI researchers are automated, ensuring they remain 'steerable, accountable, and connected to people' becomes exponentially harder. The manifesto acknowledges this, stating that 'alignment is itself a hard research problem.' The tension is clear: faster progress demands AI-driven research, but that very acceleration could outpace safety measures. For investors, this is a binary risk: either OpenAI solves alignment in time, or a catastrophic failure could trigger global regulation that halts development.

Economic Acceleration: Winners and Losers

OpenAI's second goal is to 'accelerate the economy.' The manifesto cites historical parallels: electricity increased lifespans by 23 years and incomes by 50% over the 20th century. AI, they argue, will do the same but faster. The strategic question is: who captures the value?

Winners include OpenAI (obviously), its partners like Dell, and enterprises that integrate Codex and other tools to automate software development, customer service, and data analysis. Losers include traditional software firms, consultancies, and any business model reliant on human expertise that AI can replicate. The manifesto's promise of 'widely shared' gains is aspirational; in practice, the benefits will flow to those who own the AI infrastructure and the data to train it.

For executives, the imperative is to identify which parts of their business can be augmented or replaced by AI. Those who wait risk obsolescence. The Dell partnership is a signal: Codex is coming to hybrid and on-premises environments, meaning even regulated industries like finance and healthcare can deploy AI without sending data to the cloud. This lowers the barrier to adoption but also locks enterprises into OpenAI's stack.

Personal AGI: The Ultimate Platform Lock-In

The third goal—'give everyone on Earth a personal AGI'—is the most audacious. If realized, it would be the largest platform shift since the smartphone. A personal AGI would know your preferences, habits, and goals. It would manage your schedule, negotiate bills, learn new skills, and even care for aging parents. The strategic consequence is that OpenAI would become the operating system for human life.

This is where the 'broad distribution of power' narrative becomes critical. By framing AGI as a personal assistant rather than a corporate tool, OpenAI deflects antitrust concerns. But the reality is that a single company controlling the AGI layer would have unprecedented influence over information, decisions, and economic activity. The manifesto calls for an international organization to coordinate safety, but that is years away. In the interim, OpenAI sets the rules.

For competitors, this is a direct threat. If OpenAI achieves personal AGI first, it will be nearly impossible to dislodge. Network effects, data moats, and user habits will create a winner-take-most dynamic. The only check is regulation, but the manifesto's proactive stance on safety and distribution may preempt aggressive action.

Second-Order Effects: Geopolitical and Regulatory Ripple

OpenAI's plan has clear geopolitical implications. The US-China AI race intensifies. If OpenAI succeeds, the US gains a strategic advantage. But the manifesto's call for 'national and global coordination' suggests OpenAI is preparing for a world where AI is too powerful for any single nation to control. Expect increased lobbying for international AI treaties, which could slow competitors while OpenAI builds its lead.

Regulatory risk is real. The TanStack npm supply chain attack (May 2026) exposed vulnerabilities in OpenAI's software supply chain. Security incidents will invite scrutiny. The manifesto's emphasis on safety and resilience is partly a defensive move to shape regulation in OpenAI's favor. Executives should monitor regulatory developments closely; compliance costs could rise, but early adopters of OpenAI's stack may benefit from grandfather clauses or certification programs.

Market Impact: A New Industrial Revolution

The market impact is transformative. Software development costs could drop by an order of magnitude as Codex automates coding. Scientific research accelerates, leading to faster drug discovery, materials science breakthroughs, and climate solutions. But the flip side is job displacement. The manifesto acknowledges this implicitly by stating 'entirely automating everything is not the future we want.' Yet the automated AI researcher goal suggests otherwise.

Investors should watch for signs of OpenAI's progress toward the 2028 milestone. Key indicators: hiring of alignment researchers, partnerships with cloud providers (beyond Dell), and any demonstration of AI-driven research results. If OpenAI hits its targets, the valuation of AI companies will skyrocket, and traditional tech giants will scramble to adapt.



Source: OpenAI Blog

Rate the Intelligence Signal

Intelligence FAQ

OpenAI's third phase focuses on making AGI abundant, affordable, and accessible to everyone, with three main goals: an automated AI researcher, economic acceleration, and a personal AGI for every person.

OpenAI internally believes that by March 2028, a significant fraction of its research will be done by AI systems in tandem with human researchers.

The partnership brings Codex to hybrid and on-premises enterprise environments, allowing regulated industries to deploy AI without cloud dependency, expanding OpenAI's enterprise reach.

Risks include alignment failure as AI accelerates, potential job displacement, concentration of power in OpenAI, and security vulnerabilities as seen in the TanStack supply chain attack.

Executives should evaluate integrating OpenAI's tools like Codex to gain productivity advantages, while also monitoring regulatory developments and investing in AI governance to mitigate risks.