Patch management vs. IT outages: balancing cyber risk
How AI cybersecurity threats are forcing IT teams to weigh outage risk against faster patching
After a spate of recent high-profile IT outages in just six days, including the Microsoft, FAA, and Chase outages as well as disruptions at United, Apple, and the NSA, cybersecurity teams face a hard choice in patch management: update quickly and risk downtime, or wait and risk AI-powered attacks that can turn a newly discovered vulnerability into a working exploit in minutes. Here’s how to weigh the two risks.
Key takeaways
- Recent high-profile IT outages are making teams more cautious about patch management, even as updates grow larger and more frequent.
- AI cybersecurity threats let attackers find and exploit vulnerabilities in minutes, shrinking the time available to test patches.
- Not all patches carry equal risk, so prioritize by the criticality of the system being updated.
- Effective patch management means weighing outage risk against the cybersecurity risk of leaving vulnerabilities unpatched.
Why recent IT outages make patch management riskier
While the root cause of any given outage may vary, the one thing all generally have in common is that a series of unanticipated cascading events led to an outage. As a result, IT teams are understandably going to be a lot more cautious about updating complex IT environments that are more fragile than anyone really wants to admit.
How AI cybersecurity threats speed up vulnerability exploitation
The real challenge going forward is that application environments now need to be updated more frequently than ever as cybercriminals are becoming more adept at using artificial intelligence (AI) to discover vulnerabilities and reverse engineer vulnerabilities in a matter of minutes. More challenging still, the overall size of the updates continues to increase. Any IT team running Microsoft software, for example, has taken note of the fact that patch updates have significantly increased in size as more vulnerabilities need to be fixed. Multiply the size and scope of those remediation efforts across all the applications and infrastructure that IT teams have deployed and, for the most part, they are now being asked to continuously update IT environments.
That may seem feasible on paper, but in reality most IT teams are not going to be able to make that transition overnight. Historically, IT teams were counting on the fact that it might be weeks before a vulnerability was exploited. During that time, they typically tested a patch to make sure that the cure that might take an application offline wasn’t worse than the actual disease. After all, only a small percentage of the vulnerabilities discovered have ever been exploited, largely because for one reason or another they were not easily reachable. With the aid of AI, however, it’s becoming a lot easier for cybercriminals to craft an exploit that targets a specific IT environment using novel techniques that, for instance, combine multiple low-level vulnerabilities to create a much more lethal exploit.
Not all patches carry equal risk
Ultimately, cybersecurity teams in the AI era are going to need to be able to both apply controls and other mitigations in real time to limit the scope of any potential vulnerabilities, while deploying as many patch updates as possible. The thing IT teams need to keep in mind is that not all patches and vulnerabilities are of equal risk.
While an endpoint vulnerability could be leveraged to laterally distribute malware across an entire IT environment, the impact of a patch update that results in an endpoint going offline might only create a minimal inconvenience for the business. Automatically applying updates to a client device does not represent the same level of risk as updating a mission-critical application running in a data center.
IT teams, of course, are almost by definition risk averse. Deploying patches quickly, especially if they have not had the time needed to fully test and validate them, is anathema to how they prefer to operate. This is why most updates to applications and systems are made during off hours when disruptions can be minimized. Naturally, not every application can be updated without having to reboot the IT environment, so timing is everything.
Balancing IT outage risk against cybersecurity risk
Like it or not, however, it’s become apparent that the risk that certain classes of vulnerabilities and exploits now represent to business in many cases now exceeds the potential risk of an IT outage that, if responsibly contained, can have limited impact. The challenge, as always, is determining which of those two poisons is less lethal depending on when a class of applications running where is updated.
O que os Clientes MSP precisam a seguir, da retaguarda à sala do conselho de administração
A Barracuda Research constatou que a excelência técnica continua a ser essencial, mas o crescimento futuro será impulsionado pela especialização em AI, pela resiliência cibernética e pela capacidade de comunicar o risco em termos empresariais.
Subscreva o Blogue Barracuda.
Inscreva-se para receber destaques sobre ameaças, comentários do setor e muito mais.
Relatório sobre a lacuna de competências em cibersegurança: o risco da IA lidera a lista
O inquérito da Barracuda Research a 2000 responsáveis de segurança conclui que 94% referem lacunas de competências em cibersegurança, ao mesmo tempo que a IA reduz as funções de segurança de nível inicial que ajudam a formar futuros investigadores.