Agentic AI
,
Artificial Intelligence & Machine Learning
,
Next-Generation Technologies & Secure Development
Days After Beating Musk in Court, ChatGPT Maker Moves Toward Public Debut

OpenAI is taking steps toward its IPO and could confidentially file a draft of its prospectus as early as Friday. The artificial intelligence firm is working with Goldman Sachs and Morgan Stanley to prepare the filing, according to reports.
See Also: AI Agents Introduce a New Insider Threat Model
The Wall Street Journal first reported the news on Wednesday, later confirmed independently by CNBC. The company, most recently valued at $852 billion, could go public as soon as September.
Analysts expect OpenAI’s IPO valuation to reach nearly $1 trillion, making it one of the largest debuts in corporate history.
“As part of normal governance, we regularly evaluate a range of strategic options,” an OpenAI representative said in a statement. “Our focus remains on execution.”
The IPO activity comes just two days after a federal jury dismissed Elon Musk’s lawsuit against OpenAI on technical grounds – because he failed to file his complaint within the three-year statute of limitations. The dismissal cleared the company’s path to go public without addressing Musk’s claims that it had violated its charter by launching a for-profit entity.
Reuters had previously reported that OpenAI’s IPO could come in the second half of 2026, but the quick victory over Musk may have accelerated that timeline.
The move would make OpenAI the first frontier AI startup to go public, and Elon Musk’s SpaceX and xAI are expected to file a prospectus for a $1.25 trillion IPO as soon as this week. Anthropic, whose valuation could also be in the $1 trillion range, is expected to go public later this year.
The IPO will increase the scrutiny OpenAI will face, with quarterly earnings obligations, shareholder accountability and public filings. The civil court trial showed how chaotic the company’s early years have been behind the scenes.
Earlier this month OpenAI launched Daybreak, a cybersecurity initiative that uses its GPT-5.5 models and Codex Security engine to detect vulnerabilities, validate patches and create remediation workflows. Partners using its capabilities include Cisco, CrowdStrike, Cloudflare, Oracle and Zscaler.
Unlike Anthropic’s Project Glasswing and Claude Mythos Preview model – the company’s most powerful solution for finding cybersecurity vulnerabilities, OpenAI’s Daybreak is publicly available, and customers can request a vulnerability scan on its website. Project Glasswing is only available to a select group of organizations.
