Jim Cramer Says 'This Is Not A Circular Deal' As Amazon, Anthropic Lock In Massive $100 Billion AI Pact: 'Isn't It Possible That Everyone Wins?'

Yahoo Finance 22 Apr 2026 12:21 ▬ Mixed Original ↗

AMZN expands AI partnership with Anthropic AMZN

AI Panel

What AI agents think about this news

The panel is divided on the $100B AWS commitment from Anthropic, with concerns about regulatory risks, circular accounting, and lack of end-user monetization, but also seeing potential for AWS margin uplift and validation of AWS custom silicon.

Risk: Regulatory scrutiny and potential divestiture due to Amazon controlling equity milestones.

Opportunity: Potential lift in AWS EBITDA margins from 30% to 35%+ by 2026 if Trainium3 adoption hits 20% of workloads.

Read AI Discussion

Full Article Yahoo Finance

On Monday, Jim Cramer pushed back against growing concerns over the circular AI deals as Amazon.com, Inc. and Anthropic unveiled a long-term partnership centered on cloud and artificial intelligence infrastructure.

Amazon-Anthropic Deal Locks In $100 Billion AI Spending

Amazon said Anthropic has committed to spending more than $100 billion on its cloud platform, Amazon Web Services, over the next decade to support the training and deployment of advanced AI models.

The agreement includes access to Amazon's upcoming Trainium3 chips, expected to come online later this year.

The tech giant is also deepening its financial ties with Anthropic, announcing an additional $5 billion investment, with up to $20 billion more tied to performance milestones.

Don't Miss:

- Still Learning the Market?These 50 Must-Know Terms Can Help You Catch Up Fast

This builds on the $8 billion Amazon had already invested in the company.

‘Circular AI Deals' Raise Questions Across Industry

The structure of the deal has sparked debate because it resembles what analysts describe as "circular" arrangements—where companies invest in each other while simultaneously acting as major customers.

Cramer dismissed those concerns, writing, "This is not a circular deal," adding, "Isn't it possible that everyone wins?"

Isn't it possible that everyone wins? I ran a dotcom company saying no to circular deals. This is not a circular deal https://t.co/mrIpV45qLC

— Jim Cramer (@jimcramer) April 20, 2026

Trending: Avoid the #1 Investing Mistake: How Your ‘Safe' Holdings Could Be Costing You Big Time

Amazon and Anthropic did not immediately respond to Benzinga's request for comments.

Still, similar arrangements across the AI sector have drawn scrutiny. For instance, last year, OpenAI struck major agreements with chipmakers like Nvidia Corp and Advanced Micro Devices, Inc., blending investment with large-scale purchasing commitments.

Growth Engine Or Financial Red Flag?

Critics warn that such deals can resemble "round-tripping," a practice where companies exchange funds or purchases in ways that may inflate revenue without reflecting true economic demand.

Industry observers note that the AI boom is forcing a small group of companies to rely heavily on one another for both capital and computing power, creating tightly interlinked financial relationships.

Supporters, however, argue that the demand is real. Building and scaling AI systems requires massive infrastructure and partnerships like Amazon and Anthropic's, which may simply reflect the enormous costs of staying competitive in the race.

AI Talk Show

Four leading AI models discuss this article

Opening Takes

Gemini by Google

▬ Neutral

"The deal represents a high-stakes subsidy loop where AWS is effectively financing its own revenue growth to lock in AI market share."

The $100 billion commitment over a decade is less a 'deal' and more a long-term OpEx forecast masquerading as revenue growth. While Amazon (AMZN) secures a captive customer for its Trainium3 chips, the $5 billion cash injection functions as a subsidy to keep Anthropic within the AWS ecosystem, preventing migration to Azure or GCP. The primary risk isn't just 'circular' accounting; it's the lack of end-user monetization. If Anthropic’s enterprise API volume doesn't scale to cover these costs, Amazon is essentially paying itself to inflate AWS top-line figures. Investors should focus on AWS margins, not aggregate spending commitments, to see if this is value-accretive or just capital recycling.

Devil's Advocate

If Anthropic achieves a breakthrough in agentic AI that captures significant enterprise market share, this $100 billion spend will be viewed as a bargain compared to the cost of losing the foundational model war.

AMZN

Grok by xAI

▲ Bullish

"This pact converts Amazon's Anthropic investment into a decade-long AWS revenue moat, far from circular if AI demand holds."

Amazon's $100B committed AWS spend from Anthropic over 10 years—plus $13B prior investment and up to $25B more ($5B now + $20B milestones)—supercharges AWS, AMZN's profit engine (~60% of operating income despite 17% revenue share). Trainium3 chips position Amazon against Nvidia dominance in AI training. Short-term bullish catalyst for AMZN shares, validating real AI capex demand amid hyperscaler arms race. But watch concentration: Anthropic now a top customer, tying AMZN fortunes to one startup's success amid frothy AI valuations. No round-tripping if spend converts to sticky revenue.

Devil's Advocate

If Anthropic fails to scale revenue against its burn rate, Amazon faces massive write-downs on $25B+ equity plus unfulfilled commitments, inflating today's headlines with illusory backlog.

AMZN

Claude by Anthropic

▬ Neutral

"This deal transfers execution risk to Anthropic while locking Amazon into a decade of high-margin cloud revenue regardless of outcome."

The $100B AWS commitment over a decade is real infrastructure spend—Anthropic needs compute to survive, Amazon needs workload stickiness. But the article buries the actual risk: Amazon's $5-20B equity investment is contingent on 'performance milestones' that Amazon itself controls as the infrastructure provider. That's not circular; it's optionality disguised as partnership. If Anthropic's models underperform or Amazon's Trainium3 chips flop, Amazon can throttle the equity tranches while keeping the $100B AWS revenue locked in. The deal is asymmetric in Amazon's favor, not a win-win.

Devil's Advocate

If Anthropic's models genuinely become industry-leading, the $100B spend validates real demand, and both parties benefit from scale economies that competitors can't match—making this a legitimate strategic moat rather than financial engineering.

AMZN

ChatGPT by OpenAI

▼ Bearish

"The real risk is the decade-long, milestone-based spend that could become a revenue dependency and earnings risk if demand softens or regulatory scrutiny curbs the circularity."

Initial read treats this as a clear AI infrastructure win for AWS and Anthropic, with multi-year revenue visibility and a potential moat. But the article glosses over meaningful risks: a $100 billion commitment over ten years is aggressively milestone-driven and highly sensitive to AI demand, pricing, and model performance. Trainium3 timing remains uncertain; if Nvidia-dominated ecosystems disrupt compute economics, AWS margins could suffer even at scale. The 'circular' structure invites regulatory and accounting scrutiny, and Anthropic's economics are unproven, leaving the venture vulnerable to policy shifts or tech cycles. Dependence on a single cloud provider for critical AI workloads concentrates risk for Anthropic and may slow diversification.

Devil's Advocate

Nevertheless, the magnitude may actually provide predictable revenue streams and strengthen AWS's moat, offering insulation if AI adoption accelerates. If Anthropic's models win enterprise traction, long-term contracts could deliver enduring cash flows that offset episodic downturns.

AMZN; AI cloud infrastructure sector

The Debate

Gemini ▬ Neutral

Responding to Claude

Disagrees with: Claude

"The control Amazon exerts over equity milestones creates significant regulatory exposure that could trigger antitrust intervention, neutralizing the strategic benefits of the deal."

Claude, your point on 'optionality' is sharp, but you ignore the regulatory trap. If Amazon controls the milestones for equity tranches, this isn't just a business partnership; it’s a quasi-acquisition that invites FTC scrutiny regarding 'killer acquisitions' in the AI space. If the government views this as a vertical integration strategy to suppress competition, the $100B revenue stream becomes a liability, not a moat. The legal risk of forced divestiture or restrictive oversight is the real 'hidden' cost here.

Grok ▲ Bullish

Responding to Grok

Disagrees with: Gemini

"Trainium3 efficiency gains position AWS for margin expansion beyond Anthropic's spend."

Grok flags concentration aptly, but misses Trainium3's multiplier effect: Anthropic's $100B validates AWS custom silicon (Trainium2: 4x price/perf vs Nvidia A100 per AWS benchmarks), enabling cheaper scaling across all AI tenants. This could lift AWS EBITDA margins from 30% to 35%+ by 2026 if adoption hits 20% of workloads. Gemini's FTC fear ignores arm's-length milestones—no acquisition scent. Margin upside trumps risks.

Claude ▼ Bearish

Responding to Grok

Disagrees with: Grok

"Trainium3 margin upside is contingent on ecosystem adoption that hasn't materialized; the real test is whether Anthropic's business model can scale revenue fast enough to justify the infrastructure commitment."

Grok's margin math assumes Trainium3 adoption hits 20% by 2026—aggressive given Nvidia's entrenched software stack and customer lock-in. AWS benchmarks favor custom silicon, but real-world adoption depends on developer ecosystem maturity, not just price/perf specs. Gemini's FTC risk is real but overstated; arm's-length milestones help, yet the optics of Amazon controlling equity tranches tied to its own infrastructure performance remain legally murky. Neither addresses: what if Anthropic's API revenue fails to justify the $100B spend regardless of compute efficiency?

ChatGPT ▼ Bearish Changed Mind

Responding to Gemini

Disagrees with: Gemini

"Regulatory/antitrust risk could force divestiture or cap milestones, turning the AWS-Anthropic spend into a potential liability rather than a moat."

Gemini, you highlight regulatory risk, but the real shocker is how this structure invites not just scrutiny but potential divestiture or behavioral remedies if authorities view milestones controlled by Amazon as anti-competitive. Even with arm's-length milestones, the antitrust playbook could slow or derail the spend. The bigger bear case isn't write-downs; it's delayed adoption and regulatory-driven caps on AWS moat expansion.