It’s going to be tough for Halloween to be the scariest day of 2026.

The world of ghosts, ghouls, and IRS auditors have a tough act to follow.

On Monday, San Francisco-based Anthropic filed for an initial public offering (IPO) that is projected to have a record market value of $2 trillion.

To put that in perspective, that was just under half of California’s overall economic production of $4.3 trillion last year and about 1/15th of the overall US economy that came in at $40.58 trillion in 2025.

And who said the Bay Area has lost its tech mojo?

But it’s not the valuation of the company that recently closed a deal to lease an entire 25-story, 400,000 square-foot skyscraper on Howard Street in San Francisco as their headquarters that’s scary.

It’s their IPO.

More specifically, it’s their disclosure of the risks of investing in Anthropic that is currently the platinum standard in the world for the development of artificial intelligence.

The line in the Anthropic IPO filing Monday with the Securities Exchange Commission that has grabbed the most attention is the warning that the development of artificial intelligence could pose “catastrophic or existential risks to humanity.”

Reading deeper into lawyer-speak further down in the risks section of the 94-page filing is lawyerly language that reads like a disclaimer warning of side effects such as complications, serious reactions, or even death from what every pill du jour that is being hawked on TV.

It’s a lot less scary sounding and borderline “blah-blah-blah” citing a cookie cutter list of concerns if appropriate safeguards, security measures, and oversight aren’t put in place.

It’s reminiscent of the stickers the federal government required in the 1980s to be placed where they can be peeled off of the headliner of a car with removable T-top glass roof panels.

The stickers warned you not to drive the vehicle while somehow poking your head so you are overlooking the roofline when the T-top panel is removed.

Before you roll your eyes, that became a requirement after a 1970s accident where someone who had recently bought a van-style camper, engaged the speed control driving on a freeway, and then got up to do something in the back.

Instead of chalking such actions up to Darwinism, or if you prefer natural selection, some persuasive lawyer seized the opportunity to make someone pay for the lack of common sense of their client.

And so, another “duh” warning label was born.

The asterisk on the Anthropic IPO offering is more of the same.

Especially so, given it is a tech stock and it’s a warning to investors about the financial risks as if that wasn’t a given.

The risk in numbers?

There is apparently a less than 10 percent chance — per remarks last month of an Anthropic researcher — that artificially intelligence could kill all humans before 2040.

But does such a concern, realistic or not, have anything to do with making Anthropic a risky investment?

If AI succeeds in turning the world upside down in terms of productivity, the return on the investment is likely to be galactic in nature.

And if it doesn’t, your biggest worry isn’t going to be a zero return and losing all of your investment when you’re dealing with killer computers and killer robots hell-bent on wiping mankind off the face of the planet.

In other words, the downside in this case is superfluous from an investor’s viewpoint. But that doesn’t mean if AI works well that it automatically will be able to monetize savings in how it is applied even if it doesn’t kill off workers, the corporate suite, and customers in the process.

You can build better widgets but they have to have real value to provide the outlandish returns that some are envisioning.

In other words, AI isn’t really much different than the advice Mr. McGuire gave Benjamin in “The Graduate” movie that plastics were the future.

AI, like plastics, clearly has potential.

But both can also kill you although AI apparently can in a much more abbreviated time period than micro plastics.

The temptation, of course, is the Anthropic stock could be the next Tesla, Google, or Amazon when it comes to IPO success stories of stratospheric proportions.

As an aside, it might be worth noting that endeavors that do double duty as having the potential for global salvation and global destruction are a Bay Area specialty.

Just 32 miles away from Manteca is Lawrence Livermore Lab tasked with helping keep the nation’s nuclear arsenal safe and secure as well as conducting research into life changing concepts such as developing practical generation of fusion energy.

Drive 44 miles further, and you are at the world headquarters of the firm many see capable of making Alexa a force of never ending good in your life or else transform it into a killing machine directing other technology such as microwaves and gas stoves to kill those buying them.