Coinbase Outage Exposes AWS Risk As Trading Moves Toward Recovery

Coinbase users faced degraded performance after an Amazon Web Services incident in Northern Virginia disrupted parts of the exchange’s trading stack, putting cloud infrastructure risk back at the center of crypto market reliability.

Coinbase marked its website, mobile app, and Advanced Trade services as degraded after customers were temporarily unable to transact on web and mobile. The exchange said the issue was linked to an AWS outage and later tied the disruption to increased temperatures in the affected AWS service. Coinbase also told users that funds were safe, while its team worked toward restoring trading.

The recovery process moved into a controlled restart. Coinbase said it would begin re-enabling markets shortly, with all markets placed in “Cancel Only” mode before broader trading resumed. That type of mode usually lets users cancel existing orders without opening fresh exposure, reducing execution risk while an exchange brings matching, order management, and market access back online.

AWS Health identified the affected zone as use1-az4 in the US-EAST-1 Region. AWS said increased temperatures inside a single data center impaired EC2 instances, EBS volumes, and other dependent services in the affected Availability Zone. The cloud provider shifted traffic away from the impacted zone for most services and worked to restore cooling systems before bringing affected racks back online.

No Breach Signal, But Operational Risk Is Clear

The incident was an infrastructure outage, not a confirmed security breach. Coinbase’s public updates repeatedly stated that customer funds were safe, and the available information points to cloud-service impairment rather than compromised accounts, wallets, or custody systems.

Still, the disruption matters because Coinbase is not a small app with a basic uptime problem. It is a major U.S. crypto exchange, custodian, brokerage gateway, and market-access layer for retail and institutional users. When trading access slows or stops, users lose the ability to manage exposure in real time, especially during volatile market conditions.

Reuters said Coinbase attributed its platform issues directly to the AWS outage. The same AWS event also overlapped with problems at CME Group’s CME Direct platform, although CME did not tie its own issues to AWS. That broader market context makes the outage more than a single-exchange inconvenience: one data-center thermal event affected critical financial infrastructure at the same time traders were trying to access risk systems.

For users comparing centralized exchanges, uptime, market-mode communication, withdrawal reliability, and incident transparency are part of the safety picture. Strong liquidity does not fully protect users if a platform cannot keep order access available during stress.

AWS Dependency Returns To The Spotlight

Coinbase has faced AWS-related disruption before. In a prior post-incident review, the exchange said an October AWS us-east-1 outage caused 3 hours and 17 minutes of degraded performance across services including login, trading, transfers, deposits, withdrawals, staking, onboarding, market data, and crypto sends and receives.

That history raises a sharper question for exchanges: how much operational resilience can be built when critical systems still depend on a small number of cloud regions, availability zones, and vendors. AWS Availability Zones are designed to isolate failures, but crypto exchanges need their own redundancy, routing, load shedding, failover procedures, and incident playbooks to keep user access stable when cloud infrastructure degrades.

The latest outage also comes at an uncomfortable time for Coinbase. The company recently posted a large quarterly net loss as trading revenue fell, and competition from brokerage platforms is increasing pressure on retail crypto fees. A visible outage adds another stress point, not because it changes Coinbase’s custody position by itself, but because users judge exchanges on whether they can trade, cancel, withdraw, and monitor balances when infrastructure breaks.