Investors are pushing back on sky-high AI valuations, and stock sales are getting harder

Investors are pushing back on sky-high AI valuations, and stock sales are getting harder

Bloomberg reports growing skepticism toward AI company valuations after the Firmus IPO collapse and pushback on Nscale's pricing

Share

Add us on Google
by Diego Almada Lopez

Investors are asking tougher questions about the premium valuations attached to AI and AI-adjacent companies, according to a Bloomberg report published October 19, 2026. That skepticism is making it harder for these companies to sell stock, whether through IPOs or secondary sales.

The deals that changed the mood

Firmus, an AI data center company, saw its IPO collapse within 48 hours in mid-October 2026 after failing to attract enough investor interest.

Next up for scrutiny is Nscale Ltd., which is pursuing a valuation of roughly $30 billion for its upcoming IPO. Investors are pushing back, with estimates suggesting a more realistic range of $16 billion to $19 billion.

Advertisement

The giants are not immune

OpenAI’s valuation has fluctuated between more than $500 billion and $852 billion, based on employee share sales. Anthropic, meanwhile, has eyed a potential IPO at roughly $2 trillion. Both companies now face downward adjustments to those figures, as investors weigh them against revenue trajectories and profitability metrics.

A sector-wide reassessment

Semiconductors, hyperscalers, and software names have all experienced sell-offs. The concern driving those declines is a mismatch between heavy AI spending and productivity and revenue gains that have not kept pace with that spending.

The surge in AI-related capital spending traces back to 2022 and 2023, when rapid advances in generative models fueled enormous expectations. By mid-to-late 2026, that optimism has given way to a harsh reassessment, with diminishing returns and growing debt burdens exposing fragility in investor confidence.

What this means for AI companies and investors

Nscale is the test case to watch. If it lists near its roughly $30 billion target, that would suggest the Firmus collapse was a local stumble rather than a sector-wide signal. If it prices closer to the $16 billion to $19 billion range investors have floated, that would confirm the market is repricing AI infrastructure across the board.

A potential roughly $2 trillion IPO for Anthropic would require public investors to buy into a very long growth runway. In a market demanding proof of profitability, that ask gets harder to make.

Disclosure: This article was edited by Diego Almada Lopez. For more information on how we create and review content, see our Editorial Policy.
TECHNOLOGY
Investors are pushing back on sky-high AI valuations, and stock sales are getting harder
Investors are pushing back on sky-high AI valuations, and stock sales are getting harder

Bloomberg reports growing skepticism toward AI company valuations after the Firmus IPO collapse and pushback on Nscale's pricing

by Diego Almada Lopez

Share

Add us on Google

Investors are asking tougher questions about the premium valuations attached to AI and AI-adjacent companies, according to a Bloomberg report published October 19, 2026. That skepticism is making it harder for these companies to sell stock, whether through IPOs or secondary sales.

The deals that changed the mood

Firmus, an AI data center company, saw its IPO collapse within 48 hours in mid-October 2026 after failing to attract enough investor interest.

Next up for scrutiny is Nscale Ltd., which is pursuing a valuation of roughly $30 billion for its upcoming IPO. Investors are pushing back, with estimates suggesting a more realistic range of $16 billion to $19 billion.

Advertisement

The giants are not immune

OpenAI’s valuation has fluctuated between more than $500 billion and $852 billion, based on employee share sales. Anthropic, meanwhile, has eyed a potential IPO at roughly $2 trillion. Both companies now face downward adjustments to those figures, as investors weigh them against revenue trajectories and profitability metrics.

A sector-wide reassessment

Semiconductors, hyperscalers, and software names have all experienced sell-offs. The concern driving those declines is a mismatch between heavy AI spending and productivity and revenue gains that have not kept pace with that spending.

The surge in AI-related capital spending traces back to 2022 and 2023, when rapid advances in generative models fueled enormous expectations. By mid-to-late 2026, that optimism has given way to a harsh reassessment, with diminishing returns and growing debt burdens exposing fragility in investor confidence.

What this means for AI companies and investors

Nscale is the test case to watch. If it lists near its roughly $30 billion target, that would suggest the Firmus collapse was a local stumble rather than a sector-wide signal. If it prices closer to the $16 billion to $19 billion range investors have floated, that would confirm the market is repricing AI infrastructure across the board.

A potential roughly $2 trillion IPO for Anthropic would require public investors to buy into a very long growth runway. In a market demanding proof of profitability, that ask gets harder to make.

Disclosure: This article was edited by Diego Almada Lopez. For more information on how we create and review content, see our Editorial Policy.