OpenAI’s CFO Sarah Friar confirmed that the company plans to reserve a portion of its IPO shares specifically for retail investors.
This was stated in a CNBC interview as OpenAI prepares for a potential U.S. public listing, possibly as early as the second half of 2026 with some reports eyeing late 2026 or a $1 trillion+ valuation range. Friar noted that the company tested retail participation in its most recent private funding round and saw really strong demand from individuals—reportedly oversubscribed by about 3x in some accounts. She said they will for sure extend this to the IPO.
Why This Matters
Breaking from norms: In typical tech IPOs, institutional investors like hedge funds, mutual funds, etc. get the vast majority of shares. Retail investors often receive only 5–10% and frequently miss out on the initial pop due to allocation priorities and flipping by pros. OpenAI is signaling a more inclusive approach, which could help broaden its shareholder base and align with its democratizing AI branding.
Register for Tekedia Mini-MBA edition 20 (June 8 – Sept 5, 2026).
Register for Tekedia AI in Business Masterclass.
Join Tekedia Capital Syndicate and co-invest in great global startups.
Register for Tekedia AI Lab.
Friar has tied it to building public trust in AI. The company recently raised significant private capital; one report mentioned a $122 billion round at an ~$852 billion valuation. Enterprise revenue has grown to ~40% of total, with projections it could match consumer revenue by 2026. The IPO is expected to be one of the biggest in tech history if it hits high valuations.
No specific percentage for the retail slice has been disclosed yet, and details like filing timeline, exact valuation, or how retail access will work via brokers, direct programs, or platform remain unclear. IPOs can shift based on market conditions, regulatory approvals, and internal factors.
On X, reactions are mixed: Bullish takes see it as a rare chance for Main Street to participate in a massive AI wealth-creation event. Skeptical views note it could be a way to distribute shares amid high burn rates, leadership changes, or competition from xAI, Anthropic, or a narrative play to boost hype and valuation before listing. Some compare it to past IPOs where retail enthusiasm fueled pops followed by volatility.
Historically, giving retail a meaningful allocation can generate buzz and loyalty but doesn’t guarantee strong long-term performance—plenty of hyped IPOs have underperformed after the debut. This is still early-stage planning; watch for SEC filings or further comments from OpenAI for concrete details.
Traditional tech IPOs heavily favor institutions, with retail often getting minimal allocations and missing the initial pop due to flipping by pros. OpenAI’s move—tested successfully in its recent $122B funding round where retail contributed ~$3B and oversubscribed parts by 3x—could give individuals a meaningful chance to buy at the IPO price.
If the stock debuts strongly driven by AI hype and OpenAI’s 900M weekly ChatGPT users, early retail buyers could see gains. However, many hyped IPOs experience post-listing volatility or corrections. High valuations; $852B post-money now, with IPO talk of $1T+ already bake in massive expectations around revenue growth and profitability.
Friar explicitly tied this to democratizing AI and fostering public trust—AI needs to garner trust… everyone partakes, not just a very small group. It aligns with OpenAI’s consumer-facing image versus a pure B2B play, potentially boosting user loyalty and turning customers into shareholders. Could create a more stable, long-term holder mix less prone to quick institutional exits.
It mirrors approaches like Square and Block and echoes SpaceX’s reported ~30% retail allocation. Helps signal strong demand already proven privately and supports a premium valuation. But it’s also pragmatic hygiene for becoming a public company. OpenAI is burning cash heavily; projected $14B loss in 2026 from compute and infrastructure, so the IPO provides a path beyond endless private rounds.



