Cyberthreats in 2026:
X-Force and industry experts weigh in

Cybersecurity is a contest of persistence. Adversaries probe relentlessly, defenders respond and attacks inevitably resurface—often with greater speed and sophistication enabled by advanced technologies.

At the end of last year, IBM security experts shared their predictions with IBM Think, forecasting some of the cyberthreats and solutions that are already trending in 2026–including new risks from artificial intelligence (AI), new challenges to identity management and more.

Now, in the newly released X-Force Threat Intelligence Index 2026, IBM’s X-Force team identified more patterns in how adversaries are adapting and executing their attacks in an AI- and data-focused era. We talked to IBM and industry experts about the findings in the report.

Over the past five years, major supply chain and third-party breaches increased sharply, with incidents quadrupling, according to the report. This reflects a shift in attacker behavior: rather than breaking through a single organization’s defenses, adversaries increasingly target interconnected systems and trusted integrations, such as vendors, open-source dependencies, identity integrations, CI/CD workflows and cloud interfaces.

This pattern aligns with findings from other researchers as well, including the Atlantic Council’s “Breaking Trust” report, which documented systemic weaknesses across global software supply chains and the cascading risks created by insecure components and trust relationships.

In an interview with IBM Think, Nick Bradley, Manager of IBM’s X-Force Threat Intelligence Malware team, explained why supply chain attacks are effective for adversaries.

“Attackers have figured out that they don’t need to break through your carefully guarded front door when they can walk right in through your supplier’s back door with valid credentials,” said Bradley.

Modern software, he continued, “is built on sprawling webs of dependencies, cloud services and APIs [and] the hard truth is that we’ve built highly interconnected systems without fully accounting for how this connectivity creates security vulnerabilities.”

Dr. Gregory Falco, Cornell University researcher on aerospace and semiconductor supply chain security, noted that, broadly, organizations have two options. Some organizations are responding by “vertically integrating everything” and controlling every component from end to end. Others accept that “the ecosystem will stay messy and focus instead on being ready to detect and defend as problems emerge,” he said.

In addition to the rise in supply-chain attacks, IBM X-Force observed a 44% year-over-year increase in the exploitation of public-facing applications. According to the report, public facing applications are commonly exploited due to vulnerabilities and deployment or configuration errors; this risk was amplified by supply chain attacks targeting development ecosystems and trusted infrastructure.

Recent incidents involving platforms such as Salesloft and Drift— where attackers leveraged compromised Drift OAuth tokens to access Salesforce environments—illustrate how the “compromise of a trusted third party can enable indirect access to customer environments in ways that organizations had not fully prepared for,” Christopher Caridi, Cyber Threat Analyst at X-Force Strategic Threat Analysis, told IBM Think.

“We’re moving into a world of untrusted systems,” said Bruce Schneier, a cybersecurity and public policy specialist at the Harvard Kennedy School. (Schneier was also CTO of Resilient Systems, acquired by IBM in 2016 and the originators of IBM X-Force—ed.) Solutions are scarce, he warned, and even the ones commonly suggested, such as greater transparency, only work in ideal conditions. Transparency “helps, assuming you have a customer base that is sophisticated enough to understand what they’re seeing,” he said. But he noted that many organizations still lack the visibility and expertise needed to turn that insight into meaningful defense.

Threat actors are often viewed as using highly sophisticated techniques and tactics to penetrate an organization’s defenses. But of the nearly 40,000 vulnerabilities that the X-Force team tracked in 2025, 56% could be exploited without any form of authentication.

“This highlights how often adversaries find weaknesses that do not rely on user credentials, MFA bypasses or any user interaction,” Austin Zeizel, X-Force Threat Intelligence Consultant, said to IBM Think. He added that many organizations fail to prioritize vulnerabilities that are easiest to exploit in the real world. The implication is hard to ignore; sophisticated threats exist, but many organizations are losing to adversaries who are simply taking advantage of simple, preventable gaps.

This trend also reinforces the identity versus patching debate—whether organizations should prioritize hardening identity controls or fixing exploitable vulnerabilities first.

“CISOs must treat vulnerability patching and identity hardening as parallel priorities” Caridi said. Unauthenticated flaws, he continued, demand rapid remediation to “reduce initial access risk,” while identity controls help “limit the impact when exploitation does occur.”

The growing use of AI chatbots and agents in business operations creates a new attack surface for infostealer malware. As these tools become more deeply integrated into organizational workflows, compromised systems such as AI agents containing stored chatbot credentials are an emerging risk.

In 2025, X-Force researchers found more than 300,000 ChatGPT credentials listed for sale on the dark web. Recently, OpenClaw—an open-source, locally run AI agent platform—gained notoriety as a security nightmare. “Are agents the most helpful insider threat? Of course they are,” quipped Dave McGinnis, VP and Senior Partner for global cyber threat management at IBM, on a recent Security Intelligence episode discussing OpenClaw.

Given that agents require access to data to function effectively, organizations face a difficult balancing act, McGinnis noted. Security teams are being asked to enable employees to realize the benefits of AI tools, while ensuring those capabilities are deployed safely. Striking that balance, he emphasized, is one of the most daunting challenges facing security leaders today.

Caridi echoed McGinnis’ sentiment and added that “while AI platforms themselves may become direct targets, the larger risk is the increased volume and sophistication of credential harvesting enabled by AI-assisted phishing and infostealer malware.”

Organizations that “experience fewer credential-based incidents,” he noted, are those that “consistently enforce phishing-resistant MFA and apply strong identity management practices such as conditional access, least-privilege access and continuous monitoring of authentication behavior.”

For the first time in six years, North America became the most attacked region, accounting for 29% of all X‑Force incident response cases in 2025, up from 24% in 2024. Asia Pacific, which led in prior years, dropped from 34% to 27%, marking a substantial regional redistribution of attacker focus.

Three factors may help explain why attackers have prioritized North American organizations, Ryan Anschutz, North America Leader for X‑Force Incident Response, told IBM Think:

• The region features “very high digital adoption, massive cloud footprints, and deeply connected ecosystems, all of which are moving very fast. That speed creates gaps. Add in a heavy reliance on identity, SaaS, and third-party access, and attackers simply do not need zero-days, they just need valid credentials and a little bit of patience.”
  
  

• North American organizations often sit at the center of global supply chain and customer networks, meaning a single compromise can provide downstream access to many partners and clients—a force multiplier frequently seen in incident response work.
  
  
• At the same time, improved baseline controls in parts of Asia Pacific, including stronger identity controls and network segmentation, appear to be raising the cost of entry for attackers and pushing them toward easier, higher‑return targets.

“I always say that attackers tend to follow the path of least resistance,” Anschutz added, “while still having a high return on their efforts. North America fits both, offering scale and payoff for the attacker.”

Across all regions and attack types, one theme remains constant: many security incidents stemmed from lapses in basic cybersecurity hygiene. Even as security teams adopt more automated and AI driven tools, these foundational gaps continue to create opportunities for attackers. Why do these failures persist? Zeizel attributed it to a failure to “implement foundational controls consistently at scale.”

He noted that modern environments are often “too complex for manual oversight, identity sprawl increases the impact of simple mistakes and security tools are often deployed without proper management or continuous governance.”

Zeizel emphasized that maintaining strong cybersecurity hygiene depends on several core approaches, including the following:

• Continuous, automated security practices are required to align with modern identity-centric and cloud-centric environments.
  
  
• Constant exposure management should replace periodic scanning.
  
  
• Strong identity and access controls, including least privilege access, credential protection and monitoring for misuse, are essential as attackers continue to rely on valid account abuse.
  
  
• Effective cybersecurity hygiene also depends on accurate asset inventories, consistent configuration management and the timely removal of outdated or unused systems before they become security liabilities.

Taken together, these approaches show why cybersecurity hygiene remains the first line of defense. As Falco put it, even the most advanced AI-driven protections offer little benefit “if we’re just leaving the front door open.”