IBM probably has some pretty unhappy enterprise customers right now. The company’s cloud service just suffered its fourth outage in three months, an occurrence one analyst said is an indication that structural changes are needed.
On August 11, IBM’s cloud went down for two and a half hours, with customers “experiencing failures affecting service availability, performance, and access to core cloud resources.”
Clients might have shrugged off such a quick blip had they not had to deal with similar outages in May and June as well.
IBM’s log also recorded outage incidents on May 20, June 2 and June 4, all of which featured login issues and an inability to access key services. In a report, IBM said the two incidents in June stemmed from updates made to its cloud Logs service which created a new internal traffic pattern and workload spikes that overwhelmed its Cloud Identity and Access Management (IAM) system.
David Linthicum, a cloud computing subject matter expert and former Chief Cloud Strategy Officer at Deloitte, told Fierce the frequent recurrence of these issues indicates a “foundational” issue in IBM’s architecture.
“The recurring pattern of authentication failures reveals a fragility in their foundational systems that are essential for managing key functions, including user access, service orchestration, and monitoring,” he said. “In my opinion, these disruptions highlight critical architectural weaknesses. It seems to me that their control plane is a single point of failure, and this issue needs to be addressed ASAP.”
Client concerns
To be fair, IBM isn’t the only cloud provider to have outages in 2025. Google Cloud suffered a significant outage in June that brought down major third-party internet services. Microsoft’s cloud-based Outlook services had a day-long hiccup in July, while Oracle reported power issues affecting some of its European regions in April. And AWS’ service history shows a series of blips on the radar, though these are mostly related to connectivity, error rates and latency.
But having so many similar outages in a row is not just unusual but also disruptive for clients. As a smaller player attempting to grow against hyperscale competitors, IBM can’t afford to make these kinds of mistakes and provide openings to question its credibility. In Q2 2025, it held a 2% market share in the $99 billion cloud infrastructure market, according to Synergy Research Group.
“Repeated outages are a disaster for IBM, considering that they are easily replaced by the larger hyperscalers,” Linthicum noted. “IBM may see that it loses more business from these events than the big three public cloud providers. For regulated industries such as healthcare and finance, such incidents can lead to significant compliance risks.”
He added that AI will only make reliability a more important feature for cloud customers going forward, with outages bringing the potential for “catastrophic” impacts on business clients. Why?
“You may have 100 applications bound to a single AI inference engine, and if that engine is residing on a cloud platform that can provide reliability, it stops everything,” Linthicum said.
Linthicum’s conclusion? IBM needs to tackle structural changes and engage transparently with clients if it wants to maintain any sort of trust and credibility with its enterprise customers.