Feb 17th, 2026 Market Thoughts:
AI Fears Bleed into Finance

The SaaS-pocalypse Infects Finance

I wrote about this last week, but the narrative that AI is going to fundamentally change our relationship with software has been localized mostly to pure-play tech companies. It appears that last week was the start of that narrative breaking containment a bit and it hit a sector that I wasn’t expecting, financial services. It looks like investors are starting to question business models that have exposure to the powers that AI companies are starting to build out in their models.

Let’s look at the damage:

• Charles Schwab dropped ~12%.
• Morgan Stanley shed ~5%.
• Raymond James finished Tuesday down ~8.7%—its worst drop since the March 2020 pandemic panic.

This sell off was allegedly triggered when Altruist announced an AI tool that creates personalized tax strategies by interpreting financial documents without manual entry. Suddenly, the market is panicking that we are on the verge of “vibe financial planning.” I’m skeptical. When it comes to retirement, people are inherently risk-averse. Handing the keys to your life savings to an AI agent seems a bridge too far for most. However, commoditized tax services like H&R Block and Intuit are a different beast. They are walking into an existential buzzsaw.

The tax code is effectively just a massive, convoluted rules engine, exactly the kind of environment where AI thrives. We are rapidly approaching a reality where you can simply upload your W-2s to a secure model and ask, “How do I fill this out?”

At that point, the value proposition collapses.

Why would anyone continue paying hundreds of dollars to TurboTax for a “guided experience” when the reasoning cost to process that data trends toward zero? These companies have built their empires on the friction of complexity. But if AI solves the complexity, the justification for their fees evaporates.

Maybe then, if their lobbyist war chests finally dry up, we can get the tax reform that simplifies this whole process and the entire sector can finally go away.

The Surveillance Creep: Privacy vs. Name Tag

We are witnessing a massive tension between the commercial success of consumer hardware and the steady erosion of public anonymity. Recent news from Meta and Ring suggests a strategic pivot in the Surveillance-as-a-Service economy. Companies are betting that convenience will eventually provide enough cover for features that were unthinkable a decade ago.

Meta’s smart glasses are an unexpected commercial juggernaut. The company sold over seven million pairs last year and now they are reported to be deploying their facial recognition software internally called Name Tag. This would allow wearers to identify strangers and pull up their information via an AI assistant. Zuckerberg has allegedly even internally debated whether to keep the recording light on (or if the glasses should use a different signal) when the super sensing feature is keeping a record of a person’s entire day.

Sounds like a dystopian nightmare to me. Thankfully, we’re seeing some resistance here. Ring faced a massive uproar following a Super Bowl ad for a Search Party feature that uses AI to find lost dogs. The public was quick to recognize that if you can find a dog you can use the tech to find a person. That backlash seems to have resulted in Ring killing its deal with Flock Safety which sells license plate reader tech to police. I really don’t want to live in even more of a Black Mirror episode than we already do so I hope that the public and the market continue to push back on these types of features and product developments.

Rapid Fire Observations

Some quick top level thoughts on other big things that happened last week:

The Jobs Revision Shell Game

The January jobs report showed a gain of 130,000 jobs, which looked like a win on the surface. However, the benchmark revisions for 2025 were a disaster. They wiped out 403,000 jobs that we previously thought existed, leaving 2025 with a pathetic total of only 181,000 jobs added. This creates a massive trust gap for investors. When data is this susceptible to being erased after the fact, the only pragmatic move is to assume current numbers are a fantasy.

Unity’s Crisis

Unity announced their earnings last week and took an absolute beating, dropping another 34% bringing their YTD returns to ~-58%. As I mentioned last week, this was their change to assuage concerns about AI’s impact on their business and they failed to do so.

I do however LOVE this quote from CEO Matt Bromberg regarding AI:

The impact [of AI] is gonna be much more about..the time to innovation than the time to market.

I wholeheartedly agree with that sentiment. We’ll see if Unity can help be the bridge to that future.

Discord’s IPO Cleanup

Discord is in the classic cleaning the house phase before its planned IPO. They are rolling out age verification and global teen-by-default settings to satisfy regulators. This is the standard social media lifecycle: skirt the rules to get big, then lock the doors once you need institutional buyers. While users are reportedly surging toward TeamSpeak in protest, I expect that will be shortlived. Users are already figuring out how to get around the checks and the networking effects Discord has are real.

What’s Happening Next Week

We have a few more interesting earnings report coming this week ahead of the big Nvidia earnings next week.

Figma Reporting

As I mentioned in my deep dive I am looking closely at Figma’s AI guidance and spend. The lock-up share release schedule hits on Feb 18th (putting roughly 44.4 million shares on the market). If management cannot prove their AI spend is driving real margin expansion, that supply glut will keep a lid on any price appreciation.

Klarna and the Debt Wall

Klarna’s upcoming report is the canary in the coal mine for the K-shaped economy. Consumer debt is rising. I want to see if the buy-now-pay-later model is finally hitting a wall. If the consumer is tapped out, Klarna will be the first to show the cracks in the household balance sheet.

GDP Expectations

There is a widening gap between the model and the market. Prediction markets are currently pegging Q4 GDP at 3.3 percent. However, the Atlanta Fed’s GDPNow model is forecasting 3.7 percent. If we print closer to what the prediction markets think vs. the Fed’s own predictions, I’m curious to see how the markets react. Are the big players starting to trust prediction market data more than the governments? It will be interesting to see how things play out here.

The Last Word

We’re going to continue to be on a rollercoaster this week. Sentiment isn’t shifting yet and we have a few more big earnings reports to get through and digest.

We’ll see if the AI fear continues to spread even further. I’m sure there will be another big AI advancement or new tool announcement that will give everyone an opportunity to overreact about.