Anthropic plans to proceed with an initial public offering — the first sale of a private company’s shares to public investors — later this month despite last week’s sharp selloff in artificial intelligence stocks, The Wall Street Journal reported, citing people familiar with the matter. The company, which develops the Claude AI models, is holding meetings with key investors this week as part of the pre-listing marketing effort known as a roadshow, and is targeting a Nasdaq debut next Thursday or Friday under the ticker “ANT.”

The offering has become a test of investor appetite for AI risk after last week’s market rout. Anthropic is expected to disclose its target valuation range in a filing early this week, along with the number of shares it plans to sell and the proposed price. The filing could also identify large investors that have agreed to purchase shares at the IPO price.

A week ago, investors expected Anthropic to target a valuation of about $1.5 trillion. However, some major investors have pushed for a lower figure since Thursday’s decline in AI stocks, according to the Journal. The company is now expected to accept a valuation well below the roughly $2 trillion figure discussed in recent months.

Company executives have been working with Goldman Sachs and Morgan Stanley on the valuation range. Setting the range too high could weaken demand and leave the shares vulnerable to a decline after the debut. Setting it too low could prompt criticism that Anthropic and its advisers left money on the table. Investment banks sometimes favor an opening range below initial expectations, which can later be increased if demand is strong.

A valuation range that remains near $1.5 trillion would indicate that institutional demand has held up and could support AI-linked companies listed on the Nasdaq. A materially lower range, or a decision to postpone the offering, would indicate that valuations across the sector are being reassessed. An offering of this size could also draw money from funds that hold megacap technology companies as investors make room for the new shares, potentially creating short-term pressure on those holdings.

The market backdrop has become less favorable for AI companies. Companies in the Nasdaq Composite, the technology-heavy index, lost about $2.5 trillion in value on Thursday after the US midterm elections brought a bloc of AI skeptics into office and led to the passage of moratoriums that could slow data-center construction — the physical infrastructure that houses the computing capacity AI developers rely on to train and run their models. The index has since recovered some of that loss, a development that may have supported Anthropic’s decision to continue with the IPO.

Even before the election, enthusiasm for AI stocks had been weakening amid concerns about the scale of corporate spending and the uncertainty surrounding returns. Rival OpenAI recently raised less than half of its initial funding target in its latest round, at a valuation of about $1.25 trillion, below expectations.

Anthropic could still wait for more stable market conditions. However, the Journal reported that a delay could risk triggering concern in the broader market. SpaceX’s record June IPO provides a recent benchmark: the offering raised about $75 billion at a valuation of nearly $1.8 trillion — a level close to the valuations discussed for Anthropic before the selloff.

Anthropic’s filing this week is expected to clarify the company’s proposed valuation, share count, offering price and any participation by anchor investors. In a typical US IPO, the deal is priced shortly before shares begin trading. Those details will show how closely the company’s IPO can approach the valuation investors had anticipated before last week’s selloff.