AI Is Being Blocked by People—And That’s Exactly Why Startups Will Win

In partnership with Porkbun

Level Up With a Domain Name From Porkbun

Grow your latest venture with a domain name from Porkbun. Named the #1 domain registrar 3 years in a row by USA Today, they offer hundreds of domains like .com, .tech, .shop, and others at cost — without the markup other registrars have.

Plus every domain at Porkbun comes with free WHOIS privacy, SSL certificates, Cloudflare DNS, and more. All backed by 24/7 personalized support 365 days a year. Check out Porkbun and save $1 off your domain name now.

Despite the fanfare from Fortune 500 boardrooms, AI adoption in large enterprises remains anemic. CEOs talk a big game. JPMorgan has 450 AI use cases. Yum! Brands says AI will be the “new operating system of restaurants.” But behind closed doors, very little is moving. Just 10% of U.S. companies are using AI in a meaningful way, according to recent government data.

Why? Because inside most large companies, the real friction isn’t data infrastructure or model performance. It’s institutional antibodies.

Inside the modern enterprise, in-house lawyers now outnumber the combined headcounts of many startups. Their job is to minimize risk—not accelerate innovation. With no case law to rely on, they default to no. AI models are a liability waiting to happen.

HR teams, meanwhile, have ballooned by 40% over the past decade. Their mandate is people—so when AI shows up threatening headcount, even indirectly, resistance kicks in.

Then there are middle managers, the soft tissue of the org chart. As AI starts to automate roles just below them, they begin doing the math—and quietly stall rollout. If the layer below disappears, they might be next.

You don’t need a conspiracy to explain the slowdown. Just incentives.

The friction is human. It’s political. And it’s powerful.

Even when enterprises do try to innovate with AI, they’re handcuffed by the very thing startups envy: brand equity.

Take Vogue.

In July, a Guess ad featuring an AI-generated model ran in Vogue’s print edition. It wasn’t even part of their editorial content—but the backlash was instant and fierce. Subscribers canceled. Critics declared it the “downfall of Vogue.” One viral post read: “It makes you look cheap, chintzy, lazy and desperate.”

Another AI-generated model appeared on the cover of Vogue Portugal. It didn’t matter. The response was outrage. The brand took a reputational hit—not because the tech was bad, but because the optics were uncomfortable.

That’s the risk big brands face: a legacy to protect. Every AI experiment becomes a lightning rod. A headline. A risk to the crown jewels.

Startups, by contrast, don’t have a brand to protect—they have a name to make. That’s a feature, not a bug.

Look at Artisan.

Instead of hiding behind compliance reviews and brand consultants, they plastered San Francisco with provocative billboards that read things like:

The result? Outrage. Viral tweets. Death threats. And… record-breaking sales. Over $2M in new ARR in two months.

They didn’t apologize. They amplified the controversy. They posted to Reddit, baited outrage, and converted critics into involuntary marketing allies.

It worked because startups can afford to offend the general public if they resonate with their ICP. Artisan wasn’t marketing to middle America—they were targeting AI-forward tech buyers. And they nailed it.

This is classic Innovator’s Dilemma territory.

Enterprises can see what’s coming—but they can’t react without hurting their own cash cows, disempowering key operators, or damaging their brands.

They’re too invested in how the world worked yesterday to pivot fast enough for tomorrow.

Startups aren’t.

For LPs, the gap between AI theory and AI implementation inside the enterprise is the mother of all investment signals.

This is the moment to fund companies that are:

You don’t need to bet on whether AI is real. That’s settled. The bet is on who can actually implement it without getting bogged down by lawyers, middle managers, and brand risk.

And right now, that’s startups. Not incumbents.

Eventually, some enterprises will adapt. Some will acquire their way into relevance. But the window is now—when internal resistance is still high, and AI-native startups are running laps around the red tape.

The question isn’t whether AI will change everything. It’s who gets to lead that change.

Startups don’t need permission.

And LPs who back them now won’t just profit from the transition—they’ll shape the new standard.