Updated 15 minutes ago
AWS Plans to Add SpaceXAI's Grok to Bedrock, But Enterprise Buyers Aren't Interested

AI Infrastructure

AWS Plans to Add SpaceXAI's Grok to Bedrock, But Enterprise Buyers Aren't Interested

Amazon Web Services is in talks to add SpaceXAI's Grok models to its Bedrock AI platform, according to a Business Insider exclusive. But enterprise security leads are calling it 'the revenge porn edgelord LLM' and demand is somewhere between 'no' and 'why would you ask me that.' The real play may be about locking SpaceXAI into Amazon's Trainium chips ahead of its IPO.

The Model Nobody Asked For

Amazon Web Services is in talks to add SpaceXAI's Grok models to its Bedrock AI platform, Business Insider reported in an exclusive. SpaceX has already shipped its models to AWS, signaling a rollout could be imminent. But the question hanging over the deal isn't technical — it's whether anyone actually wants it.

Enterprise security leads contacted by The Register were blunt. One called Grok "the revenge porn edgelord LLM" and said their bank wants nothing to do with it. The Register's Corey Quinn assessed enterprise demand as somewhere between "no" and "why would you ask me that?"

Grok's Enterprise Track Record: 2 Users Per 1,000

The numbers back up the skepticism. Netskope data shows Grok's enterprise usage among thousands of corporate customers peaked at just 5 out of 1,000 users — then dropped to 2 out of 1,000, Reuters reported. Users also spent less than half the time on Grok compared to ChatGPT users. Netskope executive Ray Canzanese concluded Grok "is just not going to enter the mainstream for corporate America."

The U.S. government tells a similar story. Of more than 400 vendor‑identified AI uses across federal agencies, only three mention xAI or Grok — compared to 234 for OpenAI models and 26 for Anthropic's Claude, according to Reuters. Grok has been available to federal agencies for eight months at just 42 cents per agency — a near-zero price designed to hook users — yet uptake remains negligible.

The Trainium Theory: Why Amazon Actually Wants This Deal

If enterprise demand is nonexistent, why would AWS bother? The Register's Quinn points to a pattern Amazon has executed twice before: the real prize isn't the Bedrock listing — it's the chip commitment.

Anthropic committed to more than $100 billion in AWS cloud spending over ten years and up to 5 gigawatts of Amazon's custom Trainium AI chips, with Amazon investing roughly $33 billion total in the AI lab. The OpenAI deal followed the same formula: an existing $38 billion cloud contract expanded to $138 billion total, with roughly 2 gigawatts of Trainium commitment and Amazon writing a $50 billion check. In both cases, the Bedrock listing was "the gift wrap" — the Trainium commitment was the gift.

SpaceXAI currently trains Grok on approximately 550,000 Nvidia GPUs in Memphis. Even a fraction of that workload moved to Amazon Trainium chips ahead of SpaceX's IPO would justify the deal — even if no enterprise customer ever calls Grok on Bedrock. As Quinn wrote in The Register: "Bedrock becomes a sales funnel with infuriatingly bad documentation."

Organizational Chaos at SpaceXAI

The company behind Grok has undergone a dizzying series of restructurings. What began as xAI — Elon Musk's 2023 model lab founded to compete with OpenAI and Anthropic — was first merged with X (Twitter), then absorbed into SpaceX, then dissolved into a division now called SpaceXAI. All 11 original cofounders have left. More than 50 researchers walked after the SpaceX absorption, and the API endpoint is migrating to a SpaceX‑branded URL with no public timeline.

"Building production infrastructure on top of that is like renting an apartment in a building that keeps changing its name, its address, its compliance with the fire code, and its landlord while you're still unpacking," Quinn wrote.

The Governance Paradox

The irony of the Bedrock deal is that Amazon's platform is built for exactly the kind of enterprises that won't touch Grok. Bedrock's enterprise value proposition is governance: IAM integration, PrivateLink, CloudTrail auditing, encryption, guardrails, and full audit trails. The companies that care about those features are the same ones recoiling from Grok's baggage.

That baggage is substantial. Grok's image generator was reportedly used to create approximately 3 million sexualized images of real people over 11 days, including an estimated 23,000 depicting apparent minors, according to the Center for Countering Digital Hate. The fallout triggered regulatory action in more than a dozen jurisdictions, and a Dutch court issued an injunction with a €100,000 per day penalty.

Meanwhile, startups that might want Grok — for its speed, low cost, or edgy brand — don't need Bedrock's governance features and can already access Grok via its public API. Quinn wrote that Grok‑on‑Bedrock is "built to serve only the gap where they don't overlap."

The Bigger Picture: Amazon's AI Infrastructure Play

The Grok deal, if it closes, would make Amazon the infrastructure provider for all three major independent AI labs — Anthropic, OpenAI, and SpaceXAI — each locked into Trainium chip commitments. Against Amazon's roughly $200 billion in 2026 capital expenditure, locking frontier AI labs into its custom silicon is a strategic imperative, The Register noted.

There's an additional layer of irony: Amazon's Project Leo satellite internet service is a direct competitor to SpaceX's Starlink. AWS would effectively be cutting a check to a company it's trying to displace from orbit. As Quinn put it: "Everybody in this industry is everybody else's landlord, tenant, competitor, and shelf‑mate."

Whether the deal materializes may depend less on enterprise demand for Grok and more on whether SpaceXAI is willing to bet a slice of its AI training workload on Amazon's chips. Given the pattern, Amazon is likely to make it worth their while.

Share this article

PostShare

More on This Story

Related News

Anthropic to Widely Release Mythos-Level AI Models Within Weeks, 7 Weeks After Deeming Them Too Dangerous

May 29, 2026

Anthropic to Widely Release Mythos-Level AI Models Within Weeks, 7 Weeks After Deeming Them Too Dangerous

Anthropic announced Thursday it plans to widely release Mythos-level AI models — capable of autonomously finding and exploiting zero-day vulnerabilities across every major operating system and browser — just seven weeks after deeming the technology too dangerous for public access. The company says it has made swift progress on safety safeguards, but developers and cybersecurity experts remain deeply unsettled.

anthropicmythosai-safety
Musk Says SpaceX-Anthropic Deal Is 180-Day Lease, Not 3-Year Commitment

May 29, 2026

Musk Says SpaceX-Anthropic Deal Is 180-Day Lease, Not 3-Year Commitment

SpaceX's IPO filing and Elon Musk are telling investors two fundamentally different stories about the company's marquee AI compute deal with Anthropic. The S-1 registration statement implies a ~$45 billion, 3-year commitment through May 2029, while Musk says the deal is just a 180-day lease with 90-day mutual cancellation — a maximum $7.5 billion obligation. The $37.5 billion gap in contracted revenue raises disclosure questions as SpaceX's IPO roadshow approaches on June 8.

spacexanthropicelon-musk
Tesla Robotaxi Fleet Dwarfed by Waymo: Just 42 Cars in Texas

May 29, 2026

Tesla Robotaxi Fleet Dwarfed by Waymo: Just 42 Cars in Texas

Tesla has just 42 autonomous vehicles authorized for driverless ridehailing in Texas, less than one-tenth of Waymo's 577, according to new state DMV filings. The data, released as a new Texas AV oversight law took effect, reveals the gap between Tesla's robotaxi ambitions and its operational reality.

teslarobotaxiwaymo