Why Cybersecurity Spending Alone Won't Keep Systems Running

| 2 Min Read
CISOs keep prioritizing security budgets, but recovery breaks down in handoffs between teams and providers. Here's where resilience actually fails.

Cybersecurity still tops the investment list for CISOs. Protiviti's 2027 CISO Top Risks and Opportunities study found 48% of CISOs naming it the leading strategic investment priority. The money is going out. The confidence isn't coming back. Only 41% of CISOs report high or very high confidence in their organization's ability to protect enterprise value, and just 29% feel that confident about responding with agility.

My read is that the gap isn't a funding problem. Well-run security programs still stumble when an incident hits, and the reason is usually structural. Plans and controls exist. What's missing is coordination between the people who have to act on them at the same time.

Where recovery actually breaks

Prevention gets the budget because it's measurable. But you aren't buying your way out of cyber risk. You're buying it down. Even mature organizations will eventually face an event they can't stop, often one involving a third party, a technology failure or something else outside their control.

What happens next separates the teams that recover from the teams that scramble. Security works to contain the threat while infrastructure and application teams decide which systems can safely go offline. Business leaders push to restore critical services, frequently with competing priorities. Legal weighs disclosure obligations with incomplete information. Communications drafts messages for employees, customers and regulators, often before anyone is sure what happened.

Meanwhile, a cloud, SaaS or other technology provider may control part of the environment you need to recover. That's where things usually fail. Not inside any single team, but in the handoffs and decisions between them.

In exercises like these, the hardest questions tend to sound almost too simple:

  • Who has the authority to shut down a revenue-producing system?
  • If two critical services are down, who decides which comes back first?
  • If Microsoft 365 or another primary collaboration platform is unavailable, how does the crisis team communicate?
  • If a critical provider is also down, do its recovery priorities and timelines align with yours?

If nobody can answer those in a tabletop, they won't get answered at 2 a.m. during a real outage. The useful output of an exercise isn't a reassuring score. It's the list of what broke: unclear authority, gaps in escalation paths, backups nobody could access and dependencies that teams assumed had been tested but hadn't.

Consider Quantum Dawn VIII, SIFMA's sector-wide exercise from November 2025. The polycrisis scenario included an undersea cable cut, and roughly two-thirds of participants said they had either never tested that scenario or weren't sure they had. That's a blind spot sitting in plain sight across an entire sector. Finding it in a drill costs a few hours. Finding it during a live disruption costs considerably more.

The event you plan for is rarely the event you get. So resilience can't depend on following a script. It depends on a framework that lets people collaborate quickly and make informed decisions regardless of what actually happens.

Cyber, AI and third-party risk now share the same blast radius

Risks don't arrive one at a time. A cyberattack can become a privacy issue, a regulatory matter and a business interruption within hours. A third-party outage becomes your outage, whether or not your own systems were touched. AI adds a multiplier effect, letting a single fault move faster across more data, systems and business processes.

The survey data reflects this. Cyber threats remain the highest-rated near-term concern, but privacy, third-party risk, AI, resilience and regulatory uncertainty sit closely behind. Few of those concerns live in a single department anymore.

AI doesn't change the fundamental resilience questions. It raises the stakes. An agent may read customer records, query an ERP system, act through a service account or trigger a workflow with nobody clicking a button. Start with an old question: what can it access? Then ask whose permissions or delegated authority it's using, what actions it can take, where those actions are logged and how fast you can stop it when something goes wrong.

CISOs seem to see the problem. In the 2027 Top Risks study, 51% identified improving AI governance and accountability as a leading priority, and 45% pointed to workforce readiness. Sustainable adoption starts with basic discipline: keep an inventory of agents and their owners, map the data and tools each one can reach, restrict service accounts and delegated permissions, monitor actions and make sure access can be revoked quickly.

Third parties deserve the same scrutiny. A third-party assessment can tell you a lot about a provider's controls at a given moment. It rarely tells you how that provider performs on a bad day. Most assessments still lean heavily on documentation, discussion and self-reporting. Those can confirm a security posture, but they don't prove a critical service will stay available, that the provider's recovery sequence matches yours or that people on both sides know how to disconnect, communicate and restore service together.

Start by identifying the external relationships most critical to operations. Then map which services and data depend on them, and test what happens when those relationships are interrupted. That's uncomfortable work, but it's the only way to learn whether a vendor's recovery plan actually fits yours.

This broader scope is becoming part of the CISO mandate. CSO's look at how CISOs can rise to the business resilience challenge traces the role's shift from mostly preventive work toward greater ownership of response, recovery and business resilience.

Build resilience into the modernization already underway

If cyber risk increasingly travels through interconnected systems, resilience has to be designed into how those systems get modernized. Retrofitting it later rarely works.

The shifts in the survey point the same direction. The share of CISOs selecting cybersecurity slipped from 53% to 48%, while the share selecting infrastructure modernization rose from 21% to 33%. That makes sense to me. Security and modernization are becoming the same conversation.

A cloud migration, ERP replacement, identity redesign or data-platform upgrade can retire unsupported systems and standing privileges. It can also create new concentration risk, recovery dependencies and failure points if resilience isn't built in from the start. Migrations often make environments more capable without making them easier to recover.

CISOs should use these programs to improve resilience rather than complicate it. A few questions are worth asking before the contracts are signed:

  • Are we retiring legacy platforms and connections, or just putting a new layer on top of them?
  • Are old accounts, interfaces, backup jobs and vendor dependencies actually being removed?
  • Can order-to-cash, payments, customer service and other critical processes be restored faster after the investment?
  • Are recovery, isolation and alternate-processing requirements written into the architecture and acceptance criteria from the start?
  • Have we tested the new platform's recovery path, or are we assuming modernization made it more resilient?

Closing the resilience gap

The point isn't to spend less on cybersecurity. It's to get more out of the spending. The 2027 Top Risks study shows CISOs juggling cyber threats, AI, third-party risk, resilience and modernization. On a survey, those look like separate line items. During an incident, they collide.

No organization will prevent every cyber incident. The real question is what happens when the next one gets through. Does the business know which services matter most? Does someone have clear authority to make the hard calls? Does the team know how to recover, and with which providers? If those answers only emerge during the crisis, the resilience gap is still open.

Closing it will take less new technology than most leaders expect and more rehearsed decision-making. The CISOs who make that shift first will be the ones whose organizations recover quickly when the next disruption arrives.

Source: Robert Martinez · www.csoonline.com

Comments

Please sign in to comment.