OpenAI CEO Forecasts Superintelligence by 2028 at AI Impact Summit

In a defining moment at the AI Impact Summit 2026 in New Delhi, OpenAI CEO Sam Altman delivered a startling prediction that has sent ripples through the global technology community. Speaking to an audience of policymakers, industry leaders, and researchers, Altman suggested that early forms of superintelligence could emerge within the next few years, specifically targeting 2028 as a potential milestone where the world's aggregated computational intelligence might surpass human capacity.

The summit, held at the Bharat Mandapam, has become a pivotal platform for global AI discourse, marking the first time such a high-level AI governance gathering has been hosted in the Global South. Altman’s comments underscore a rapid acceleration in AI development that is outpacing even aggressive historical forecasts.

The Shift of Intellectual Capital to Data Centers

One of the most profound metrics Altman introduced was the changing geography of intelligence. "By the end of 2028, more of the world's intellectual capacity could reside inside data centers than outside them," Altman stated. This visualization frames the imminent future not just as a technological upgrade, but as a fundamental shift in where the planet's cognitive processing power is located.

He emphasized that this transition is driven by the exponential scaling of compute infrastructure. The progression from systems struggling with high school mathematics to those capable of deriving novel theoretical physics results has occurred in less than a decade. Altman framed this as a "generational challenge," comparing the rapid buildup of AI infrastructure to the scaffolding of previous industrial revolutions, but with a much steeper vertical trajectory.

Divergent Timelines: Altman vs. Hassabis

While the summit featured a broad consensus on the transformative power of AI, there were nuanced differences in the timelines and risk assessments provided by leading figures. Google DeepMind CEO Demis Hassabis, who also addressed the summit, offered a slightly more conservative but equally urgent timeline.

Comparison of Key Predictions at AI Impact Summit 2026

Hassabis warned that while we are at a "threshold moment," current systems still lack the consistency and long-term planning capabilities of human cognition. However, he cautioned that the arrival of Artificial General Intelligence (AGI) is "on the horizon," likely by the end of the decade.

Critical Risk Warnings: Biosecurity and "Bad Actors"

The optimism surrounding scientific breakthroughs—such as AI's potential to cure diseases or solve fusion physics—was balanced by stark warnings regarding safety. Demis Hassabis was particularly vocal about the dual-use nature of advanced AI systems.

Hassabis highlighted two immediate areas of concern that require urgent attention:

• Biosecurity: The potential for AI models to assist in the creation of novel pathogens or biological weapons.
• Cybersecurity: The risk of AI systems being used to exploit digital vulnerabilities at a scale and speed humans cannot defend against.

"We need to worry about things like bio and cyber risk in AI very soon," Hassabis urged, noting that "bad actors"—ranging from rogue individuals to nation-states—could repurpose these powerful tools for harmful ends. He advocated for a rigorous "scientific method" approach to AI safety, where guardrails are built and tested with the same precision as the models themselves.

Three Pillars for a Democratic AI Future

In his address, Altman outlined a three-part framework designed to guide the responsible development of superintelligence. He argued that without these pillars, the benefits of AI would not be equitably distributed.

1. Democratization of AI: Altman rejected the notion of "effective totalitarianism" as a trade-off for safety. He insisted that the only path forward is one that increases human agency and liberty, ensuring broad access to AI tools rather than concentrating power in the hands of a few labs or governments.
2. Resilience Through Society-Wide Defense: Acknowledging that no single lab can secure the future, Altman called for a "society-wide approach" to defense. This involves creating systems that can withstand misuse, ensuring that defensive AI capabilities always outstrip offensive ones.
3. Co-evolution with Society: He emphasized that AI development cannot happen in a vacuum. "Most of the important discoveries happen when technology and society meet," Altman noted. He championed an iterative deployment strategy, allowing society to adapt to and shape the technology in real-time rather than being presented with a finished, overwhelming product.

India’s Strategic Role in the AI Landscape

A recurring theme throughout the summit was India's unique position to influence the trajectory of global AI. With one in 100 million people in India already using ChatGPT weekly, and a third of those being students, the country is rapidly becoming a testing ground for mass AI adoption.

Key Factors Positioning India as an AI Powerhouse:

• Demographic Scale: A vast, young population that is digitally native.
• Adoption Rates: India is currently the fastest-growing market for Codex, OpenAI’s coding tool.
• Governance Leadership: As the world's largest democracy, India is seen as a critical counterweight to authoritarian models of AI governance.

"India is well-positioned to lead in AI -- not just to build it, but to shape it and decide what our future is going to look like," Altman remarked. This sentiment was echoed by Prime Minister Narendra Modi, who pitched a "develop in India, develop for the world" vision, emphasizing ethical and inclusive AI.

Economic Disruption and the "GPU Standard"

Addressing the economic anxieties that often accompany AI advancements, Altman was candid about the disruptions ahead. He noted that in many specific tasks, "it'll be very hard to outwork a GPU," signaling a fundamental change in the labor market. However, he maintained an optimistic outlook on human adaptability, suggesting that technology always displaces jobs while simultaneously creating "new and better things to do."

The consensus from the summit is clear: the era of theoretical discussions about superintelligence is ending. The technology is transitioning into a tangible, high-stakes reality. With leaders like Altman and Hassabis predicting pivotal breakthroughs before the decade is out, the focus has shifted entirely to preparedness, governance, and ensuring that the intelligence residing in data centers serves the humanity outside of them.