AI-driven patching surge linked to global tech outages, experts say
US officials and cybersecurity experts attribute recent global tech outages to the surge in AI-driven vulnerability discovery and emergency patching. The rapid deployment of updates is causing instability in banking and travel systems.
WASHINGTON - Increased efforts to rapidly patch newly discovered software vulnerabilities in the age of artificial intelligence are likely a major factor behind recent disruptions to how people work, access bank accounts and travel, according to cybersecurity experts and US officials. While the precise causes of the recent outages remain unclear, the surge in patching activity is viewed by security professionals as a significant contributing element to the instability.
US Secretary of State Marco Rubio, speaking to reporters in Manila this week, said that company patching efforts are, at times, "creating short-term disruption in the way the system works." Rubio’s comments provide a high-level official acknowledgement of the link between security maintenance and operational downtime. The statement was made in the context of ongoing discussions about the resilience of digital infrastructure in a rapidly changing technological landscape.
Patches are software and operating system updates intended to fix security vulnerabilities found within a program or product. The advent of easily accessible AI tools and large language models has supercharged the discovery of these vulnerabilities, fundamentally changing the pace at which security flaws are identified and must be addressed. This acceleration has shifted the burden from finding flaws to managing the sheer volume of fixes required.
The scale of this acceleration is evident in recent data from major technology providers. Microsoft this month has addressed more than 1,200 software vulnerabilities, a figure that is nearly 20 times higher than in the same month a year ago, according to cybersecurity firm Qualys. This exponential increase in the volume of patches required has placed significant strain on the infrastructure of companies relying on these systems. The data suggests that the rate of vulnerability discovery is outpacing the traditional capacity of organizations to test and deploy updates safely.
Yuval Itzchakov, chief technology officer and co-founder of autonomous offensive security company A Security, noted that AI has made finding critical vulnerabilities fast and cheap. Consequently, the number of flaws that demand immediate patching has grown in parallel with the discovery rate. Itzchakov explained that emergency patches often go into core infrastructure within hours, frequently without the full testing cycle. He described these patches as intrusive by design, as they change runtimes, kernels and network components that everything else depends on. "That speed and scale of change is itself a source of failure," Itzchakov said. The lack of comprehensive testing in emergency scenarios increases the risk of introducing new bugs or compatibility issues into critical systems.
The broader industry context further illustrates the pressure on defenders. Josh Bressers, director of coordinated vulnerability disclosure at cybersecurity firm VulnCheck, stated that in 2026 there have been 100,000 publicly disclosed vulnerabilities. These are tracked by the industry as Common Vulnerabilities and Exposures, or CVEs. Of the 100,000 publicly disclosed vulnerabilities in 2026, 1,363 were known to have been exploited in the real world, according to Bressers. While the majority of disclosed flaws are not actively exploited, the existence of a significant number of exploited vulnerabilities underscores the urgency for defenders to prioritize their efforts.
Bressers emphasized that understanding which vulnerabilities need attention first is critical given the tight resources available to defenders. "Not all vulnerabilities are created equal," Bressers wrote in a LinkedIn post. The challenge for security teams is to triage the massive influx of CVEs to identify those that pose the most immediate threat to their specific infrastructure. This triage process is complicated by the speed at which new vulnerabilities are disclosed and the limited bandwidth of security operations centers.
Anne Neuberger, general partner at Andreessen Horowitz, offered a perspective on the dual nature of AI in cybersecurity. She said that while AI is helping attackers find and weaponise vulnerabilities, it is a more "valuable friend" to those tasked with defending key computer networks. Neuberger, who served as deputy national security adviser under former president Joe Biden and led incident response to serious national incidents, said that AI is enabling enterprises to find 10 times more vulnerabilities. However, she cautioned that when patching at such a scale, some of these updates can cause problems. Her comments highlight the paradox of AI-driven security: the same technology that accelerates defense also accelerates the complexity of maintenance.
The impact of these disruptions has been felt in the financial sector, where at least one major bank was rushing to address a wave of recent patches. A source at the bank, speaking on condition of anonymity, said there is a lot of work needed to manage the influx of security updates. The banking sector is particularly vulnerable to such disruptions because its systems are highly interconnected and require high availability. Even short-term outages can have significant economic consequences and erode customer trust.
Despite the operational challenges, some financial analysts suggest the long-term impact on consumer trust may be limited. Ronald Gestiehr, director of wealth management at Luttner Financial, said that technology has made banking "incredibly convenient." He expressed the view that while occasional outages can be frustrating, they are unlikely to have a significant impact on consumer confidence or change how most people manage their money. Gestiehr added that financial institutions have invested heavily in technology and security, and that occasional disruptions are an unavoidable consequence of maintaining robust systems. "What matters most is how quickly those issues are identified, resolved and communicated to customers," Gestiehr said. This perspective suggests that the market may be resilient to the temporary instability caused by patching surges.
The situation highlights a growing tension in the cybersecurity landscape: the need for immediate remediation of AI-discovered flaws versus the risk of introducing new instability through rapid, large-scale system changes. As AI tools become more accessible, the volume of vulnerabilities will likely continue to rise, requiring ongoing adaptation from both defenders and the infrastructure they protect. Organizations must balance the imperative to secure their systems against the operational risks associated with frequent, untested updates.
US officials and industry leaders are monitoring the situation closely, acknowledging that while AI enhances defensive capabilities, the sheer volume of required patches presents a new category of operational risk. The recent outages serve as a case study in the unintended consequences of accelerating security responses in a digital environment that is increasingly reliant on complex, interconnected systems. The incident underscores the need for better coordination between software vendors, infrastructure providers and end-users to manage the transition to AI-driven security practices.
The dialogue between government officials and private sector experts underscores the importance of balancing speed with stability. As the AI age progresses, the definition of a secure system may need to evolve to account for the dynamic nature of vulnerability discovery and the potential for patch-induced failures. Traditional models of security maintenance, which assumed a slower pace of vulnerability disclosure, are no longer sufficient to address the current threat landscape.
For now, the focus remains on mitigating the immediate disruptions while developing more robust frameworks for managing the influx of AI-driven security updates. The experience of the past few weeks is expected to inform future policies and practices in both the public and private sectors regarding software maintenance and security patching. Experts are calling for improved testing protocols, better communication channels for patch deployment and greater investment in automated security operations to handle the increased volume of vulnerabilities.
The reliance on AI for vulnerability discovery is not expected to slow down. On the contrary, as large language models and other AI tools become more sophisticated, the rate of discovery is likely to increase further. This will place even greater pressure on the infrastructure that supports the global digital economy. The challenge for the industry is to harness the benefits of AI for security without compromising the stability of the systems that billions of people depend on for their daily work, financial transactions and travel.
The comments from US Secretary of State Marco Rubio in Manila signal that the issue has moved beyond the technical sphere into the realm of national security and economic stability. The disruption of banking and travel systems has implications for international commerce and the free flow of information. As such, the response to these outages is likely to involve not just technical fixes but also policy discussions about the standards and practices for software maintenance in the AI era.
Cybersecurity firms like Qualys, A Security and VulnCheck are playing a crucial role in providing data and analysis to help organizations navigate this new reality. Their findings highlight the unprecedented scale of the vulnerability landscape and the need for a more nuanced approach to patch management. The industry is moving towards a model where security is a continuous process rather than a periodic task, but this shift is proving challenging to implement without causing collateral damage to system availability.
The case of the major bank rushing to address patches illustrates the practical difficulties faced by organizations in the private sector. While they have the resources to invest in security, they also have a duty to their customers to maintain service availability. The tension between these two obligations is at the heart of the current crisis. Resolving this tension will require innovation in both software development and operational practices, as well as a shared understanding of the risks involved in rapid patching.
The role of AI in this context is multifaceted. It is a tool for discovery, a driver of urgency and a source of complexity. As organizations continue to integrate AI into their security operations, they must be aware of the potential for these tools to create new vulnerabilities through the speed and scale of their interventions. The goal is to use AI to enhance security without undermining the stability of the digital infrastructure that supports modern society.
The recent outages are a wake-up call for the global tech industry. They demonstrate that the pace of technological change is outstripping the ability of organizations to adapt their processes and infrastructure. As the AI age advances, the need for collaboration between governments, industry and academia will only grow. The lessons learned from these incidents will be critical in shaping the future of cybersecurity and ensuring that the benefits of AI are realized without compromising the reliability of essential digital services.
Elected News
-
Bangladesh Seeks UN Women Support on Digital Harassment and Women’s Empowerment
-
Bangladesh, Australia Discuss Human Trafficking, Irregular Migration and Rohingya Crisis
-
Bangladesh Plans Internet-Free Access to Mobile Banking and Financial Apps
-
DU to Open Undergraduate Admission Applications on 11 November, Tests Begin 5 December