Security Leaders Cited Trust Barriers to AI Adoption

Experts identified persistent risks in AI-generated code and gaps in core data protection at a recent summit.

Updated on Oct. 2, 2026 in Cybersecurity

Isometric editorial illustration of a heavy industrial steel server rack and a vault door, representing security infrastructure in an enterprise data environment. — Security leaders at the CyberRisk Alliance Summit report that 92% of organizations identify persistent trust barriers and vulnerability risks as obstacles to AI adoption. AI Illustration. Upload story photo >

Live Poll

Do you trust organizations to safely manage your personal data when using AI tools?

Security leaders at the CyberRisk Alliance Cybersecurity Summit reported a significant trust gap in AI, with 92% citing barriers to adoption. Despite high confidence in AI for alert triage, only 35% of organizations have actively deployed such tools.

Why it matters

The security industry is weighing the efficiency of autonomous triage against the risks of AI-generated vulnerabilities, such as reintroduced SQL injection flaws. Organizations remain cautious, prioritizing human accountability and transparency over black-box automation.

While 97% of security leaders express confidence in AI-driven alert triage, 92% report a trust barrier holding back broader adoption. Currently, 72% of leaders are comfortable delegating only low-to-medium severity alerts to autonomous systems.

The players

CyberRisk Alliance

An organization providing intelligence and events for the cybersecurity industry.

CISA

The Cybersecurity and Infrastructure Security Agency that leads national efforts to understand and manage cyber risk.

The details

AI models trained on public code repositories can inadvertently propagate legacy security flaws like SQL injection—the unauthorized manipulation of database queries—and cross-site scripting. To mitigate these risks, experts recommend network segmentation to limit the blast radius of compromised credentials and continuous authorization to prevent lateral movement by non-human agents. CISA incident response data highlights that many organizations lack essential immutable or off-site backups, leaving them vulnerable to automated exploitation.

October 1, 2026: The CyberRisk Alliance Cybersecurity Summit was held in Bellevue, Washington.
Next 12 to 18 months: Organizations are advised to evaluate if new security capabilities measurably mitigate specific risks.

The Tech Race

These findings on backup and segmentation gaps follow a pattern established by CISA incident response protocols. The industry currently sits at a junction where autonomous triage must be reconciled with the foundational security controls necessary to prevent large-scale compromise.

Security teams should prioritize the implementation of immutable backups and network segmentation as immediate defenses against AI-powered threats. Over the next 18 months, leaders must focus on verifying that any adopted AI triage tools measurably improve response times without introducing new code vulnerabilities.

The takeaway

The security industry is shifting toward a model where AI is used for triage while humans retain control over final outcomes. Watch for future benchmarks in the next 18 months to see if specific AI tools can be proven to mitigate risks rather than introduce new attack vectors.