Just for kicks, I went digging through Jersey Mike’s IPO documents—because sometimes the most revealing parts of a public offering aren’t the glossy investor slides. They’re the dense, unglamorous sections where companies are forced to describe risks, dependencies, and future plans in plain regulatory language.
And like many people, I started with an assumption that felt almost intuitive: a sandwich chain—one with a footprint built on locations, franchise relationships, supply chains, and operational execution—wouldn’t have much reason to name-drop “AI” in a way that stands out. Not in the way you’d expect from a software company, a cloud platform, or a company whose core product is explicitly tied to machine learning.
But the documents do reference AI-related topics. The interesting part isn’t that “AI exists” (it does). It’s how quickly the term has become a default reference point in corporate filings—how it shows up not necessarily because it’s central to the business model, but because it has become part of the modern risk-and-future-technology vocabulary. In other words: the AI bubble isn’t staying confined to tech companies. It’s migrating into mainstream corporate language, and IPO filings are one of the clearest places to watch that migration happen.
What follows isn’t a claim that Jersey Mike’s is building an AI product. It’s a closer look at what it means when a company that sells sandwiches feels compelled to discuss AI in its public-market narrative—and what that says about the current state of hype, expectations, and corporate incentives.
A filing is not a press release
IPO documents are written for regulators and investors, not for customers. That matters, because the tone and structure are different. A press release can be selective; a filing has to anticipate questions. It has to cover contingencies. It has to explain what could go wrong, what could change, and what the company believes about the future.
So when AI appears in that context, it often lands in categories like:
1) Risk factors: concerns about technology disruption, cybersecurity, data handling, competitive dynamics, or operational impacts.
2) Future technology expectations: statements about how the company might adopt new tools, improve efficiency, or evolve customer experiences.
3) General business context: broader language about industry trends and the environment in which the company operates.
Even if AI isn’t a core revenue driver, it can still be relevant as a “background variable” that affects everything from marketing effectiveness to fraud prevention to supply chain optimization to employee productivity tools. And once a term becomes a shorthand for a whole class of technologies, it becomes easy to include it—even when the company’s actual implementation is modest or uncertain.
That’s where the hype signal comes in. Not because the company is lying, but because the inclusion itself reflects how pervasive the concept has become. It’s the difference between “we use AI in a meaningful way” and “AI is part of the landscape we must acknowledge.” Those are not the same thing, but they can look similar to readers who skim.
The AI term as corporate insurance
One of the most underappreciated dynamics in modern corporate communications is that companies don’t just describe what they do—they also describe what they might do, and what they might need to respond to.
In that sense, AI functions like corporate insurance. If you’re raising capital and entering public markets, you want to demonstrate that you understand the environment you’re stepping into. You want to show that you’re aware of technological shifts that could affect your operations, your costs, your competitive position, or your compliance obligations.
If AI is widely discussed across industries, then failing to mention it can create a different kind of risk: the risk of appearing naive. Investors may interpret silence as lack of awareness. Regulators may interpret silence as incomplete disclosure of relevant trends. Competitors may already be positioning themselves as “AI-enabled,” even if their actual use cases are limited.
So companies add AI language—not always because they have a breakthrough, but because they’re trying to keep pace with the narrative that markets now treat as baseline.
This is why the AI hype has become so sticky. It’s not only that companies want to attract attention. It’s that they want to avoid being penalized for not participating in the story.
The sandwich shop effect: when AI becomes a universal modifier
Jersey Mike’s is not alone. Many non-tech businesses have begun to incorporate AI into their public-facing language. But there’s something uniquely revealing about seeing it in a filing from a company whose day-to-day reality is fundamentally physical: ingredients, stores, labor, franchise operations, logistics, and customer demand.
When AI shows up here, it suggests that the term has moved beyond a specific technology category and into a universal modifier—something that can be attached to almost any modern business process.
Consider the kinds of areas where AI could plausibly matter to a restaurant chain, even if it’s not the headline:
– Demand forecasting: predicting sales by location, time of day, seasonality, and local events.
– Inventory and supply planning: reducing waste and improving availability.
– Pricing and promotions: optimizing offers based on customer behavior patterns.
– Fraud detection: monitoring payment anomalies, loyalty program abuse, or account takeovers.
– Customer service automation: chat-based support, order issue resolution, or personalization.
– Marketing optimization: targeting campaigns more effectively and measuring outcomes.
– Security and compliance: detecting suspicious activity, improving monitoring, and managing data risks.
None of these require a company to be “an AI company.” They require tools that can be described as AI, even if the underlying systems are relatively conventional by tech standards. The problem is that the word “AI” can cover everything from sophisticated machine learning models to basic automation and rules-based systems that marketers label as AI.
So when a filing references AI, the reader has to ask: is this a real capability, a planned initiative, or a general acknowledgment of technological change? The documents may not always make that distinction obvious to casual readers. And that ambiguity is part of what makes the hype feel out of control.
The hype isn’t just in the tech sector—it’s in the expectations
The most important shift isn’t that AI is mentioned outside tech. It’s that markets now treat AI as a proxy for future competitiveness.
In the early days of the AI boom, the narrative was: “AI will transform everything.” That was vague but exciting. Over time, it became more specific in investor minds: “AI will determine winners and losers.” Once that belief takes hold, companies feel pressure to demonstrate alignment with the future.
An IPO is a moment when a company is essentially asking the public to believe in its trajectory. That belief is shaped by what the company signals about its ability to adapt. If AI is the dominant adaptation story, then mentioning AI becomes a way to reassure investors that the company isn’t stuck in the past.
But reassurance can become performative. When every company mentions AI, the term loses precision. It becomes less about what’s actually happening and more about whether the company is “in the conversation.”
That’s why the Jersey Mike’s example resonates. It’s not that a sandwich chain shouldn’t mention AI. It’s that the mention reflects how thoroughly the AI narrative has colonized corporate language.
Risk factors: the quiet place where hype hides
Risk factor sections are often where the most telling language appears, because they reveal what management thinks could derail the business.
AI-related risks can be legitimate and serious. For example:
– Increased cybersecurity threats: AI can be used by attackers, and defenders may need new tools.
– Data privacy and governance challenges: more data processing can increase exposure.
– Operational disruption: new technologies can fail, be misused, or create compliance burdens.
– Competitive pressure: competitors adopting AI could reduce costs or improve customer experiences.
– Regulatory uncertainty: AI-related rules may evolve, affecting compliance requirements.
These are real issues. But there’s also a pattern: once AI becomes a standard topic, risk sections can start to read like a checklist of modern anxieties. That doesn’t mean the risks are fake. It means the framing can become generic.
Generic framing is where hype creeps in. If AI is mentioned as a broad risk without clear linkage to concrete systems, it can function as a rhetorical placeholder. It signals awareness without committing to specifics.
For investors, that creates a challenge: how do you evaluate whether AI is a meaningful driver of performance or merely a line item in the narrative?
The answer is in the details—and those details are often harder to find than the word “AI” itself.
The “default reference point” problem
When a term becomes a default reference point, it changes how people read corporate documents.
Instead of asking, “What does this company do with AI?” readers start asking, “Why did they mention AI at all?” That question is a symptom of a larger issue: the market has trained itself to expect AI everywhere.
Once that expectation exists, companies face a dilemma:
– Mention AI and risk sounding like you’re chasing hype.
– Don’t mention AI and risk looking behind the curve.
Most companies choose the first option, because the second option can be punished by perception. Even if the company’s actual AI usage is limited, the mention can help it avoid being categorized as outdated.
This is how hype spreads without requiring a corresponding increase in capability. The language moves faster than the technology.
And IPO filings are a perfect stage for this because they’re designed to be comprehensive. They’re also designed to be comparable across companies. If AI is a common element in filings, it becomes easier to include it—even if the company’s situation is unique.
What “accurate” AI disclosure looks like
There’s a healthy version of AI disclosure, and it’s worth stating clearly.
Accurate AI disclosure would typically include:
– Specific use cases: what processes are affected, what outcomes are targeted.
– Implementation maturity: whether AI is deployed, piloted, or planned.
– Governance and controls: how data is handled, how models are monitored, how errors are managed.
– Measurable impact: cost savings, improved conversion, reduced fraud