Gartner has identified deepfakes, AI application breaches, prompt injection, and software supply chain attacks as key cyber threats that enterprises should watch out for in 2026–2027. The diagnosis is that a multi-layered defense system is needed rather than a single security measure, as the proliferation of generative AI increases the risk of identity impersonation, AI system manipulation, and development environment breaches.

Gartner announced in a press release on the 9th that it evaluated these threats based on 'threat signals' and 'organizational response capabilities.' Threat signals refer to the quantity and quality of available information, while organizational response capabilities refer to the level at which the threat can be effectively managed.

Deepfake has been identified as a field with a growing potential for use as a means of identity impersonation as voice, video, and image generation technologies become more sophisticated. Gartner explained that real-time deepfakes could be exploited for biometric authentication procedures, online meetings, social engineering attacks targeting employees, and recruitment processes.

John Watts, a VP analyst at Gartner, stated that the ways attackers utilize deepfakes continue to evolve, making it difficult to respond with a single security measure alone. He recommended that companies pursue the strengthening of business processes, raising security awareness, and adopting deepfake detection technologies together.

AI application breaches have also been classified as a major threat. This is because the attack surface is expanding as companies operate in-house AI agents, third-party integration tools, and AI apps exclusively for employees. AI applications with weak security can lead to the exposure of sensitive data or credentials.

Gartner identified prompt injection as a threat targeting LLM-based AI systems. The explanation states that if an attacker inputs crafted commands to distort model behavior, it can lead to the leakage of sensitive information, the execution of unauthorized tasks, and the bypassing of security controls. Input validation, AI security testing, runtime guardrails, and anomaly monitoring were suggested as countermeasures.

Software supply chain threats were addressed as an issue of growing importance due to the proliferation of open source software and AI-based development environments. Gartner recommended that companies build software asset inventories and use verified repositories for third-party code, container images, and AI models.

It also explained that vendors must be required to submit SBOMs and AIBOMs, and that CI/CD pipeline protection, build artifact signing, and least privilege-based access controls must be applied. Gartner stated that as AI adoption expands, security leaders must identify threat signals and adjust their response frameworks.