TD SYNNEX 发布云安全与身份治理科普文章，无任何商业或财务增量事实

How organizations can strengthen visibility, governance and resilience as cloud and AI adoption accelerate.

A cloud bill spikes unexpectedly. A privileged account remains active long after it should have been removed. An AI application gains access to information it was never intended to use.

They may look like separate issues, but they often have the same root cause: organizations lost track of who has access to what.

Protecting cloud infrastructure is still important. But cloud security now extends far beyond servers and networks. It includes identities, applications, data, AI workloads and the business operations that depend on them. As a result, security is becoming a continuous business capability rather than a periodic technology initiative.

Cloud fraud can reveal broader security weaknesses.

Unexpected cloud consumption is often viewed as a financial issue. However, abnormal spending patterns can sometimes indicate a deeper security problem.

Weak access controls, exposed credentials and unauthorized activity may all contribute to unexpected resource usage. Left unaddressed, those same weaknesses can increase the risk of data exposure, operational disruption or broader compromise.

Compromised accounts, exposed API keys, excessive permissions and poorly governed resources can all contribute to cloud abuse.

Cloud fraud is often treated as a billing problem when it may actually be an early warning sign of a security issue.

Budget thresholds, consumption monitoring and anomaly alerts can also help identify unusual activity before costs or risks escalate.

Cloud environments are growing faster than many organizations can track.

Resources can be created in minutes. Access can change just as quickly. New integrations and services often appear faster than governance processes can adapt.

Research from Orca Security’s 2025 State of Cloud Security Report found that 32% of cloud assets were in a neglected state, highlighting the ongoing challenge of maintaining oversight at scale.

Effective cloud security starts with maintaining an accurate inventory of resources, understanding who or what can access them and monitoring changes as they occur.

Those fundamentals often reveal risks that would otherwise go unnoticed.

Access is becoming the primary battleground.

The most valuable asset in a cloud environment may no longer be the infrastructure itself, the identities trusted to access it.

Today’s attackers often do not need to bypass security controls. Gaining access through a trusted account, an over-permissioned service identity or an exposed credential can be enough.

One common challenge is that organizations tend to review employee access more often than application access. Yet applications, integrations, service accounts and automated workflows frequently retain permissions long after they are needed.

Reducing risk requires regular review of user permissions, privileged roles, service accounts, application access, API keys, credentials and dormant accounts.

The fewer permissions an account has, the less damage it can cause if it is compromised. Strong authentication, multi-factor authentication (MFA), least-privilege access and continuous identity monitoring remain among the most effective cloud security controls available.

AI is introducing a new identity challenge.

AI systems now do far more than generate content.

Increasingly, AI agents can access data, connect to applications, call APIs and perform actions across business systems. In many cases, they function less like software tools and more like operational actors working inside the environment.

AI deployments often introduce new cloud resources, data stores, APIs, identities and high-performance computing workloads. Each expands what organizations need to monitor and govern.

Organizations are now managing autonomous systems alongside people and applications.

According to Microsoft’s Digital Defense Report , 88% of enterprises are experimenting with AI agents, highlighting how quickly these technologies are becoming part of day-to-day operations.

Before deploying an AI system, organizations should establish data-access boundaries, document ownership, understand system integrations and define a process for removing access when it is no longer needed.

Many organizations move quickly to deploy AI tools before deciding who owns them, how they will be monitored or how access will be removed. The same controls used for users and applications should also be applied to AI-enabled systems from the beginning.

The biggest cloud security mistake may be assuming the work is finished.

The moment a security control is implemented, the environment around it begins to change.

New users join the organization. Applications evolve. Integrations are added. Permissions accumulate over time.

A permission that made sense six months ago may no longer make sense today.

Regular access reviews, configuration assessments and alert testing are some of the simplest ways to identify issues before they become incidents.

Security policies are most effective when they are translated into controls that are consistently monitored, tested and enforced.

Effective guardrails can help prevent insecure deployments, excessive access, exposed credentials and uncontrolled consumption before they become incidents.

Prevention alone is no longer enough.

Perfect security remains unattainable. What separates resilient organizations is how quickly they can identify and contain emerging threats.

Speed is becoming one of the most important security controls.

Microsoft’s Digital Defense Report found that exposed cloud workloads were attacked after an average of 5.3 hours, underscoring how quickly organizations must respond.

The goal is not perfect security. The goal is reducing the time between detection and action.

Organizations that can detect suspicious activity quickly, escalate concerns efficiently and respond decisively are often better positioned to limit disruption.

Turning strategy into action.

Strong security programs are built on consistent operational practices:

Enforcing MFA and conditional access

Applying least-privilege access and conducting entitlement reviews

Securing API keys, secrets and credentials

Establishing configuration baselines and policy enforcement

Monitoring budgets, consumption and anomaly alerts

Reviewing AI services and integrations regularly

Documenting incident response and escalation procedures

Conducting customer verification and enhanced due diligence when appropriate

The challenge facing organizations today goes beyond protecting more technology to managing the growing number of users, applications, services and AI agents trusted to access business resources.

Organizations best positioned to thrive will not be those chasing every new threat. They will be the ones that build security into how they operate, making access reviews, monitoring and response part of everyday business decisions.

Follow TD SYNNEX on LinkedIn for more cybersecurity insights, practical guidance and resources throughout Cybersecurity Awareness Month.

5 minutes read
The post In the cloud, identity may matter more than infrastructure. appeared first on News .