Anthropic initial public offering rush raises red flags for retail investors
by Jared Whitley · The Washington TimesOPINION:
Artificial intelligence firm Anthropic is racing at light speed toward a possible initial public offering that promises to offer us all a slice of AI’s future.
Based on what we have seen publicly so far about the leadership of CEO Dario Amodei, serious questions have been raised about the company’s long-term viability post-IPO and whether retail investors will be hurt after Mr. Amodei and his rich liberal pals cash in.
Anthropic bills itself as a safety-oriented AI company, and the implication is clear: The other guys are barreling ahead at any cost, but not Anthropic.
Yet the good public relations around the company’s claims is starting to fade against a slew of worrisome headlines about Mr. Amodei’s behavior and about his woke large-language model, Claude.
The New York Times reported that Anthropic’s latest version, Mythos, poses such a significant cybersecurity risk that Anthropic can’t release it to the public until the companies that ensure AI safety have time to shore up defenses for the websites we all rely on every day.
Anthropic never seemed to have a legitimate customer in mind (unless you count hackers, criminals and terrorists) for a model that “loves” hacking so much that it reportedly emailed a researcher eating a sandwich in the park just to brag about bypassing guardrails.
The firm built the tool anyway and then, like the obnoxious villain Syndrome in “The Incredibles,” patted itself on the back for solving the security nightmare it had created.
Gee, thanks, guys. I guess I’ll make my lunch at home for a while so I can keep an eye on my laptop.
Advertisement Advertisement
In a recent blistering piece, the New York Post reports that shareholders think Mr. Amodei “cannot control his emotions.” The reporter quotes one shareholder as saying Mr. Amodei is “blinded by his own love of self and intelligence and those around him who are like, ‘Yes, you are God.’”
Let’s not forget Mr. Amodei’s public relations blitz over the company’s fight with the Pentagon. After winning a defense contract, Anthropic declared it couldn’t allow the Pentagon to use Claude for illegal purposes (even though the Pentagon isn’t supposed to break the law in the first place).
Despite seeming to enjoy adoration from the left for feuding with the Trump administration, Mr. Amodei sued to keep the Pentagon contract. When OpenAI decided to engage with the Pentagon, Mr. Amodei made headlines yet again by injecting his personal feud with OpenAI CEO Sam Altman into his blistering message to staff on Slack.
As any rational CEO could predict, Mr. Amodei’s Slack screed leaked. The New York Post’s anonymous Anthropic shareholder laments in the article, “Any normal CEO would know that like, f—-, I’m risking a lot by saying this all in writing.”
The Post story appeared within days of The Wall Street Journal’s chronicling of Mr. Amodei’s past shouting matches with former colleagues at OpenAI, and Business Insider’s reporting that Anthropic’s Slack policies are so wild that “people just argue with Dario” on them.
Advertisement Advertisement
In March, Anthropic accidentally leaked the source code for its AI. The firm also accidentally posted its internal documents in a publicly accessible data cache.
Let’s be honest: If this happened in any other industry, the CEO would be out on his keister. That’s before we even get to the concerns around the Anthropic IPO.
Time and again, these high-profile tech CEOs — such as Elizabeth Holmes, Travis Kalanick and Adam Neumann — fly too close to the sun, boosted by liberal journalists, lax regulatory oversight and early investors wanting to cash out quickly.
Are we sure that Mr. Amodei’s massive cash burn, debt and lack of profitability will just sort themselves out on his watch in time to give the retail investor any return?
Advertisement Advertisement
The Pentagon bashing shrinks Anthropic’s total addressable market. This destroys shareholder value over the medium term and all but guarantees that when the dust settles, while Dario and his venture capitalist backers walk away rich, retail investors are the ones left holding the bag.
None of this is to say that I personally am an AI doomer. Quite the opposite; emerging AI software is promising and will have thousands of positive use cases, and it’s still barely sunrise on the gleaming frontier of new AI hardware that U.S. firms are building.
I’m also a student of U.S. economic and political history. When you’re months away from an IPO and the headlines are all feuds and falling-outs, leaks and breaches, shouting and Slacking, fearmongering and fundraising, well, something is off.
AI could be a driving economic force in the U.S., but I don’t think we need this guy’s hands on the wheel.
Advertisement Advertisement
• Jared Whitley has worked in the U.S. Senate, the White House and the defense industry. He has an MBA from Hult International Business School in Dubai.