The incident began at approximately 5:25 PM PDT on May 7 when temperatures rose inside availability zone use1-az4, part of AWS's US-EAST-1 region in Northern Virginia. The heat triggered a power loss that damaged hardware racks, knocking out eight AWS services and cascading into platform failures across financial markets. AWS said it brought additional cooling capacity online by 10:11 PM PDT on May 8, putting the total recovery window at more than 17 hours from the initial detection to stabilisation.

Parametrix Insurance, which monitors cloud provider availability, reported that approximately 30 percent of all tracked providers experienced disruptions of varying severity during the event.

Coinbase, which operates its exchange on AWS infrastructure by design, cycled through sequential trading modes as the outage unfolded: normal operations gave way to Cancel Only mode, then an Auction phase, before the platform began a gradual restoration. The exchange told users that customer funds were safe throughout and directed them to the AWS Health Dashboard for status updates.

CME Group simultaneously reported that users could not access CME Direct, its trading login portal. CME later said it had "completed the necessary technical maintenance" to restore access but did not confirm any link to the AWS event, and Verse Press is treating that connection as unconfirmed.

---

The timing compounded an already difficult news cycle for Coinbase. The company released its first-quarter 2026 earnings on the same day the outage struck. Revenue came in at $1.41 billion, below the $1.52 billion analyst consensus, and the exchange posted a GAAP net loss of $394.1 million. Despite those figures, Coinbase recorded a historically high crypto trading market share of 8.6 percent and derivatives volume of approximately $4.2 billion, representing 169 percent growth year over year.

One day before the outage, Coinbase, AWS, and Stripe had jointly announced the integration of Coinbase's x402 discovery layer and USDC wallet infrastructure into Amazon Bedrock AgentCore Payments, a platform allowing AI agents to transact in stablecoins.

The gap between that announcement and the subsequent infrastructure failure drew immediate comment from technology commentator Gergely Orosz. His criticism carried particular force because Coinbase's CEO had recently commented publicly about non-technical staff shipping production code, making the infrastructure failure a focal point of that specific line of scrutiny.

---

The outage is the second major AWS US-EAST-1 incident to knock Coinbase offline within seven months. In October 2025, a DynamoDB DNS race condition in the same region caused a roughly 15-hour outage estimated to have generated $2.8 billion in economic losses across affected services, including Binance and KuCoin.

StatusGator, a service monitoring firm, has documented US-EAST-1 as Amazon's most frequently disrupted region, a pattern Amazon attributes to the region's operational scale rather than structural problems.

The November 2025 CME outage, which was linked to a separate cooling failure at colocation provider CyrusOne, halted FX, equity index, rates, gold, and crude oil futures during Asian trading hours.

No on-chain volume spikes or confirmed price dislocations tied directly to the May 2026 outage window have been independently verified at the time of publication. Traders blocked from Coinbase would have had access to alternative centralised exchanges or decentralised platforms such as Uniswap, which operates entirely on-chain and was unaffected.

---

The disruption carries concrete implications outside the United States. In India, where exporters and importers routinely hedge commodity exposure through CME futures on crude oil and gold, even a temporary halt in price feed availability introduces ambiguity into physical trade settlement. The Multi Commodity Exchange of India references global benchmark prices for its own contracts, so disrupted CME data ripples into domestic markets. For Indian fintech firms exploring USDC payment rails for cross-border remittances, the fragility of the AWS layer connecting Coinbase's stablecoin infrastructure is a direct operational risk.

Across Sub-Saharan Africa, where Coinbase has been expanding retail access in markets including Nigeria, Ghana, Kenya, and South Africa and promoting USDC as a dollar proxy in high-inflation economies, users who rely on the platform as a store of value would have encountered the same access blackout experienced globally.

African developers building on Coinbase's Base layer-2 network and using AWS-hosted APIs with US-EAST-1 dependencies would have encountered the same tooling disruptions during the outage window.

Binance, the dominant exchange across the region, runs on geographically distributed infrastructure and was not affected during this event. Operators weighing diversification toward Binance should note that the platform navigated significant regulatory pressure in Nigeria during 2024 and 2025, a regional risk factor that sits alongside its infrastructure advantages.

---

For developers and fintech operators building payment or trading infrastructure on cloud-native financial rails, the incident points to a design gap that two US-EAST-1 outages in seven months has made harder to ignore. Single-region cloud dependencies, absent circuit breakers or multi-cloud failover, carry measurable downtime risk.

At the time of publication, neither AWS nor Coinbase had made public statements addressing whether architectural or resilience changes are planned following the second major disruption within the same region in seven months.

Whether regulators in markets such as India, where both the Securities and Exchange Board of India and the Reserve Bank of India hold active oversight mandates over exchanges and payment infrastructure, will treat repeated exchange outages as a disclosure or operational continuity concern remains an open question heading into the second half of 2026.