Daily cloud and web hosting news coverage by HostingDiscussion.com
Today’s Storystream

Nvidia plans to invest $2 billion in AI cloud provider Nebius as the two companies deepen collaboration on hyperscale infrastructure for emerging AI workloads. The partnership focuses on building large-scale “AI factories” and deploying Nvidia’s next-generation accelerated computing platforms across Nebius data centers in the US and Europe. Nebius aims to expand its AI cloud capacity beyond 5GW by 2030 to support growing demand for inference and agentic AI systems.

Attackers are increasingly breaching cloud environments by exploiting vulnerabilities in third-party software rather than relying on weak passwords, according to Google’s latest Cloud Threat Horizons report. Software flaws now account for 44.5% of initial access incidents, surpassing credential-based attacks for the first time. Researchers also observed faster exploitation cycles and growing abuse of SaaS integrations and OAuth tokens, exposing new risks across connected cloud ecosystems.

Security researchers warn that poorly configured Salesforce Experience Cloud environments are being exploited to extract sensitive data through guest accounts. According to BleepingComputer, attackers linked to the ShinyHunters group are targeting the platform’s /s/sfsites/aura API using a modified AuraInspector tool. By abusing excessive guest permissions, the group claims it accessed data from hundreds of organizations, highlighting persistent risks tied to cloud configuration errors.

Enterprise investment in public cloud services is on track to exceed $1 trillion in 2026, according to IDC, as organizations intensify spending on AI platforms and cloud-native infrastructure. While SaaS remains the largest share of the market, Platform-as-a-Service is expanding fastest as companies build data-heavy AI applications and analytics systems. Banking, retail, and software sectors are leading adoption as cloud infrastructure becomes central to modern IT strategies.

AI cloud provider CoreWeave has introduced a flexible pricing framework designed to better match the uneven demand patterns of AI workloads. The model adds Flex Reservations and Spot instances alongside traditional options, allowing enterprises to balance guaranteed GPU capacity with lower-cost, interruptible compute. The approach targets a growing challenge in AI infrastructure—unpredictable inference traffic—while giving teams more control over GPU spending and workload scaling.

Cloud marketplaces are streamlining enterprise software purchasing, but experts warn the convenience may obscure deeper operational risks. According to Forbes, Technology Council members note that one-click procurement often outpaces governance, leaving gaps in integration planning, security reviews, and cost oversight. Licensing mismatches, autoscaling infrastructure charges, and uncoordinated team adoption can inflate long-term spending, highlighting the need for stricter governance before organizations commit to marketplace software.