The digital landscape is continuously evolving, and with it, the security challenges organizations face are also changing. A recent readership poll highlights that agentic AI risk is at the forefront of cybersecurity priorities as we head into 2026.

The poll, influenced by discussions among cybersecurity experts, aimed to gather insights on the most pressing security issues. The findings reveal that nearly half of the respondents believe agentic AI will be the primary attack vector for cybercriminals and nation-state threats by the end of 2026. This sentiment reflects a growing concern as enterprises increasingly adopt agentic AI for various applications, from streamlining operations to enhancing competitive advantages in software development.

Agentic AI: A New Target for Cybercrime

With the rise of agentic AI, security professionals express concern over the potential for increased vulnerabilities. Rik Turner, chief analyst for cybersecurity, emphasizes the risks associated with the expanded attack surface due to the autonomy and access levels of AI agents. He warns that a rush to adopt agentic AI could lead to the deployment of insecure code, thus exacerbating security vulnerabilities.

Moreover, the proliferation of open-source AI agents and shadow AI—applications used by employees without oversight—further complicates the security landscape. Omdia's internal surveys support the poll findings, indicating that AI adoption is a major concern for corporate security teams.

Deepfakes: An Emerging Threat

Another significant concern is the rise of deepfakes, with 29% of respondents indicating that they are becoming a favored tactic among cybercriminals targeting high-profile individuals and organizations. Deepfakes have gained traction, particularly in state-sponsored campaigns, and many businesses are not adequately investing in defenses against these threats.

Melinda Marks, a cybersecurity director at Omdia, notes that while prevention is essential, the focus should also be on rapid detection and response. This dual approach is vital as organizations strive to protect themselves against increasingly sophisticated attacks.

Cyber Risk Elevated to a Tier 1 Priority

Despite varying predictions, only 13% of respondents believe that cyber-risk will be recognized as a Tier 1 operational priority by corporate boards in 2026. Turner expresses skepticism regarding whether current trends will significantly shift boardroom priorities toward cybersecurity. However, the rise of agentic AI is expected to influence how cybersecurity risks are perceived within enterprises.

Amy Worley from BRG suggests that recognizing agentic AI as a critical security risk could prompt boards to implement necessary safety measures. She cautions that as AI systems gain autonomy, the potential for errors or malicious actions increases substantially, necessitating proactive measures from executives and boards.

Password Elimination and Passkey Adoption

Only 10% of respondents anticipate that password elimination and passkey adoption will become standard practice in 2026. Although password security remains a concern, more robust authentication methods are not receiving the investment they require. Adam Etherington from Omdia highlights the risks posed by the prevalence of agentic systems and the need for enhanced security measures amidst these emerging threats.

While some progress has been made in adopting passkeys, Turner echoes the sentiment that a complete transition away from passwords is unlikely in the near future. The debate continues over whether passwords will fade into obscurity or remain a staple of cybersecurity practices.

In conclusion, as we move toward 2026, the landscape of cybersecurity is set to change dramatically with the rise of agentic AI. Organizations must prioritize security to address the evolving threats posed by AI technologies, deepfakes, and the overall increase in cyber risks. This shift presents an opportunity for security leaders to align with business goals and ensure robust protective measures are in place.

Source: Dark Reading News