Core Automation Targets Massive Funding for Continual-Learning AI

Core Automation, a newly launched artificial intelligence startup founded by former OpenAI Vice President of Research Jerry Tworek, is reportedly in talks to raise between $500 million and $1 billion, signaling bold ambitions just weeks after its inception. The prospective funding round underscores strong investor confidence in AI approaches that aim to move beyond the limitations of today’s static large language models.

Tworek brings deep technical credibility to the venture. During his seven-year tenure at OpenAI, he played a key role in foundational advances such as the Chinchilla scaling laws, early coding models including Codex and ChatGPT, reinforcement learning applied to robotics, and advanced reasoning systems like o1 and o3 that led industry benchmarks. His departure from OpenAI in January 2026 followed remarks that certain long-horizon, high-risk research directions were difficult to pursue within large institutional structures.

At the heart of Core Automation’s vision is continual learning. Rather than relying on periodic retraining cycles, the company aims to build AI systems that learn and adapt continuously from real-world use. These self-improving, agent-based models are targeted at enterprise automation, software engineering, and robotics—domains where responsiveness, context awareness, and long-term adaptation are essential.

The scale of the proposed funding implies a valuation exceeding $5 billion, positioning Core Automation among the most aggressively backed AI startups at launch. Such capital would enable access to massive compute resources and support the development of frontier-scale models. Industry observers expect interest from top-tier venture firms familiar with Tworek’s track record.

While the approach promises to address persistent challenges such as model staleness and hallucinations, it also faces intense competition for compute, talent, and infrastructure. Core Automation’s trajectory will serve as a test case for whether individual AI leaders can still outpace established labs—and whether continual learning represents the next major leap in intelligent automation.