Everyone’s talking about Anthropic’s Super Bowl ad and Sam Altman’s meltdown response. But the real story isn’t what happened during the Super Bowl—it’s what happened the week before. And the month before that. And the year before that.

Let me walk you through what’s actually going on here.

The Ad That Broke Sam Altman

During the Super Bowl, Anthropic ran an ad with a simple message: “Ads are coming to AI. But not to Claude.”

Sam Altman’s response? A novella-sized Twitter rant calling Anthropic’s ads “clearly dishonest” and “deceptive.” He called the company “authoritarian” and accused them of being on a “dark path.”

Over a 30-second ad about not showing ads.

If that seems like an overreaction, you’re not wrong. But to understand why Altman is so worked up, you need to see what he’s not tweeting about.

Let’s Talk About What’s Actually Dishonest

Here’s the timeline Sam Altman doesn’t want you focused on.

Early 2024: The Setup

At the start of 2024, OpenAI was 15 times larger than Anthropic. ChatGPT was a household name. OpenAI had the brand, the funding, the mindshare. Anthropic was the scrappy competitor founded by ex-OpenAI employees who left over—ironically—concerns about safety and leadership.

The gap seemed insurmountable.

June 2024: The First Cracks

OpenAI released GPT-4o on May 13, 2024. It was a big deal—multimodal, faster, cheaper.

Five weeks later, Anthropic released Claude 3.5 Sonnet. And it didn’t just compete with GPT-4o. It beat it. On coding benchmarks, on reasoning, on the things developers actually care about.

The response window: 5 weeks. Still measured in weeks, but competitive.

October 2024: Anthropic Takes the Lead

On October 22, 2024, Anthropic released something OpenAI didn’t have: Computer Use. Claude became the first frontier AI model that could actually control your computer—move the cursor, click buttons, type text.

It was a genuine first-mover advantage. And it outperformed OpenAI’s o1 model on coding benchmarks: 49% on SWE-bench (a benchmark that measures AI’s ability to resolve real GitHub issues) versus 41%.

For the first time, Anthropic wasn’t just competing. They were leading in a category that mattered: developer tools.

OpenAI’s response? Radio silence. They’d release their version, “Operator,” three months later in January 2025. Not hours, not days. Months.

November 2024: The Arms Race Accelerates

Between November 12-24, 2024, all three major AI labs—OpenAI, Anthropic, and Google—released flagship models within a 12-day window.

One tech outlet put it perfectly: “Each one landed in the market faster than most legal tech vendors deploy minor updates.”

The competitive timing was getting tighter. Weeks had become days.

February 2025: Claude Code Changes Everything

Anthropic launched Claude Code as an agentic command-line tool in February 2025. By May, it was generally available.

Within six months, it hit $1 billion ARR.

That’s faster than ChatGPT’s initial growth—one of the most successful product launches in tech history.

And it wasn’t just revenue—developers genuinely preferred it. Survey after survey showed the same thing: for coding, Claude was the choice. 78% developer preference. “Claude smokes GPT4 for Python and it isn’t even close,” as one developer put it.

By mid-2025, Claude Code was responsible for 4% of all public GitHub commits. Think about that. Four percent of all code being written publicly was coming from one tool that didn’t exist a year earlier.

August 2025: Caught Red-Handed

Here’s where it gets interesting.

In August 2025, Anthropic revoked OpenAI’s API access to Claude.

Why? Because OpenAI’s developers were using Claude. Not casually, not for research. They were integrating Claude directly into their internal testing infrastructure ahead of the GPT-5 launch.

Anthropic’s spokesperson put it bluntly: “Claude Code has become the go-to choice for coders everywhere, and so it was no surprise to learn OpenAI’s own technical staff were also using our coding tools ahead of the launch of GPT-5. Unfortunately, this is a direct violation of our terms of service.”

OpenAI’s response tried to frame it as “industry standard benchmarking.” But there’s a difference between evaluating a competitor’s product and plugging it into your internal development tools.

The irony is perfect: the company claiming to lead in AI had its own developers preferring the competitor’s product.

December 24-25, 2025: The Holiday “Gift” Fiasco

On Christmas Eve, Anthropic announced they were doubling usage limits for Claude users through the end of the year. A nice holiday gesture.

Within hours—not days, hours—OpenAI announced the exact same thing. Double the usage limits through January 1st. Same timeframe, same offer, same framing as a “holiday thank you.”

I got the OpenAI email on December 27th. It felt disingenuous because it was disingenuous. When Claude Code showed that message in my terminal about increased limits, it felt thoughtful. When OpenAI’s email landed two days later with an identical offer, it felt like they were scrambling not to look bad.

Multiple tech outlets noted the suspicious timing: “Both companies framed it as a holiday thank-you. Both picked the same week. Both addressed the same pain point. That’s not a coincidence.”

The response window had shrunk from weeks to days to hours.

February 5, 2026: The 15-Minute War

Here’s the part that really matters.

On February 5, 2026, both OpenAI and Anthropic released competing flagship coding models. Claude Opus 4.6 and GPT-5.3 Codex.

The time between releases? 15 minutes.

But here’s what most coverage missed: according to multiple industry sources, both companies had originally planned to release these models simultaneously at 10:00 AM PST. A coordinated launch.

Anthropic released at 9:45 AM—15 minutes early.

OpenAI responded at 10:00 AM—exactly when they’d originally planned, now 15 minutes after Anthropic.

Someone broke the gentleman’s agreement. And the 15-minute response window proves OpenAI was watching, ready to react the moment Anthropic moved.

That’s not strategy. That’s obsession.

The Numbers Don’t Lie

While all this competitive theater was playing out, the market was shifting underneath OpenAI’s feet.

Early 2024: OpenAI was 15x larger than Anthropic End of 2024: The gap had shrunk to 5x

Anthropic’s revenue growth in 2025:

  • December 2024: $1B ARR
  • June 2025: $4B ARR (quadrupled in six months)
  • Approaching $7B by end of 2025

In the AI coding market specifically—the category that drives developer mindshare and enterprise adoption—Anthropic wasn’t just competing. They were winning:

  • 78% developer preference for Claude over GPT for coding tasks
  • Claude Code hit $1B ARR faster than ChatGPT did
  • 4% of all GitHub commits from Claude Code
  • Consistent benchmark superiority on coding tasks

And here’s the kicker: this is a $34 billion market growing at 17-27% annually. Losing in coding doesn’t just hurt OpenAI’s pride. It threatens their strategic position as developers are the influential early adopters who drive enterprise decisions.

What Altman Is Really Mad About

So let’s come back to that Super Bowl ad.

Anthropic’s claim was simple: they won’t show ads in Claude. Is that controversial? Is it “clearly dishonest”?

No. It’s a business model statement. It’s verifiable. They either show ads or they don’t.

But Altman called it deceptive and unleashed a Twitter storm that made him look unhinged.

Why?

Because by the time that ad aired, Anthropic had:

  • Shrunk OpenAI’s market lead from 15x to 5x in one year
  • Become the dominant player in AI coding
  • Forced OpenAI into reactive, copycat behavior
  • Caught OpenAI’s own developers using their product
  • Pushed response times from weeks to literal minutes

The Super Bowl ad wasn’t the problem. It was the final indignity.

OpenAI is the household name. ChatGPT is synonymous with AI for most people. But in the one category that matters most for technical credibility—coding—they’re getting beaten. Badly.

And the man who presents himself as a visionary leader of the AI revolution is spending his time on Twitter having a meltdown over a 30-second ad while his company plays catch-up.

The Pattern of Reactive Behavior

This isn’t just about product releases. It’s about a pattern:

Sam Altman’s public image: Safety-focused philanthropist, advocate for AI regulation, thoughtful leader steering humanity toward beneficial AI.

The documented record:

  • Testified to Congress about the need for AI regulation, then lobbied the EU to water down the AI Act
  • Presents OpenAI as an innovation leader while copying competitor moves within hours
  • Called out for not being “consistently candid” by his own board (who fired him in November 2023)
  • Had a co-founder testify under oath about his “consistent pattern of lying” (Ilya Sutskever, October 2025)
  • Claims “industry standard benchmarking” when caught using competitor’s product to build his own

There’s a word for saying one thing and doing another. It’s the same word Altman used for Anthropic’s ad: deceptive.

Why This Matters

Companies compete—that’s capitalism. But the February 5th coordinated launch that became a 15-minute competitive response? That’s not normal competition.

These companies are in private communication about timing, but publicly positioning as fierce rivals. They’re coordinating launches one day and scrambling to undercut each other the next.

That’s not healthy competition driving innovation. That’s two companies so obsessed with each other that they’ve forgotten to focus on what they’re actually building.

And when your own developers prefer the competitor’s product? When you’re caught using their tools to build your own? When you’re copying their holiday promotions within hours?

That’s not leadership. That’s desperation.

The Real Story

The Super Bowl ad controversy is a distraction. It’s Sam Altman trying to control the narrative while the ground shifts beneath him.

The real story is in the timeline:

  • Weeks (June 2024) → Days (November 2024) → Hours (December 2025) → Minutes (February 2026)

It’s a company that was 15x larger than its competitor shrinking to 5x in twelve months.

It’s developers voting with their keyboards, choosing Claude for the work that matters.

It’s the emperor having a Twitter meltdown because someone pointed out his clothes.

Anthropic’s Super Bowl ad wasn’t dishonest. But Altman’s response? That told the truth more clearly than any advertisement could.

The research for this post included analysis of official company statements, benchmark data, market reports, and documentation of product releases from both companies. All timeline events are sourced from public announcements and verified through multiple outlets.

Update 2026-02-10: OpenAI declined to comment on the coordinated launch timeline. Anthropic provided a statement reiterating their commitment to “building AI systems people can trust” but did not address the February 5th timing specifically.