Sixty-five percent of IT leaders say their cybersecurity defenses cannot withstand attacks powered by artificial intelligence, according to new research from Lenovo.
The survey, part of the company’s third Work Reborn report, found that only 31 percent of IT leaders feel confident in defending against AI-driven threats. While AI has boosted business efficiency, it has also fueled a new class of cyberattacks that many enterprises are unprepared to counter.
Lenovo’s report highlights growing concern over the sophistication of AI-enabled crime. Generative AI tools are allowing attackers to craft phishing schemes, deepfake impersonations, and polymorphic malware that mutate to evade detection.
Such threats can spread across cloud services, endpoints, applications, and data repositories faster than traditional defenses can adapt.
Beyond external attacks, insider risks are also on the rise. Seventy percent of IT leaders surveyed said employee misuse of AI is a major concern. More than 60 percent identified AI “agents” themselves as potential insider threats.
The report also noted that AI models, training data, and prompts have become high-value targets for manipulation, raising the stakes for organizations that rely on the technology.
Industry analysts have issued similar warnings. Gartner projects that by 2027, 90 percent of successful AI use in cybersecurity will involve tactical automation rather than replacing human roles. McKinsey has cautioned that companies relying on conventional defenses risk falling behind as threats evolve.
Experts say addressing these risks requires embedding AI into defensive systems. Current adoption is hampered by legacy infrastructure, talent shortages, and budget constraints, leaving organizations vulnerable as threats outpace protections.
The report comes as businesses worldwide struggle with balancing AI’s promise and its perils. While enterprises are eager to adopt AI tools for productivity and cost savings, many are finding their security operations ill-equipped for the new era of cybercrime.
Lenovo said enterprises should move toward “AI-native defenses” that adapt in real time. The company is pushing for the integration of intelligence into devices and platforms to detect and respond to attacks more quickly.
Analysts say such changes may be unavoidable. As attackers exploit AI’s ability to mimic legitimate activity and shift tactics instantly, businesses without adaptive defenses could face escalating breaches.
The findings underscore what security professionals have long warned: traditional defenses are no match for adversaries armed with AI.