Enterprises Lag in Defending Against AI-Powered Threats
According to PwC’s 2027 Global Digital Trust Insights report, which surveyed nearly 4,000 business and technology leaders across more than 70 countries, organizations are struggling to keep pace with the speed of AI-driven cyber threats. The report identifies attacks targeting enterprise AI systems as the top concern leaders feel least prepared to address. Specifically, compromise by autonomous botnets (53%), adversarial attacks (52%), and data poisoning (52%) are cited as the primary AI-related threats for which defenses are inadequate.
Prompt injection attacks, recognized as the #1 threat in the current OWASP Top Ten for LLM applications, remain a critical vulnerability. As noted by OpenAI in late 2025, prompt injection — akin to social engineering in its persistence — is unlikely to ever be fully mitigated, posing ongoing risks of data leakage or unauthorized command execution.
AI as Defense Faces Adoption Barriers
While AI is widely regarded as the #1 cybersecurity tool for threat detection and alerting, trust in fully autonomous AI agents remains low. Only 22% of leaders would permit “fully autonomous execution by AI agents for cyber defense without human approval.” The main barriers include concerns over reliability and maturity (55%) and insufficient workforce skills in AI oversight and governance (44%).
However, the report argues that only autonomous defensive AI can match the speed, scale, and sophistication of AI-assisted attacks. Relying on human-in-the-loop systems may create dangerous delays, even as full autonomy introduces new accountability challenges and expands the attack surface.
Quantum Threat Preparedness Remains Critically Low
Looking ahead to quantum risks, the report highlights the growing danger of “harvest now, decrypt later” strategies, where attackers exfiltrate encrypted data today in anticipation of future decryption via quantum computers using Shor’s algorithm — a scenario referred to as ‘Q-day.’ Despite this looming threat, only 21% of survey respondents reported implementing quantum-resistant cryptographic measures.
The report stresses that while current encryption can be upgraded to quantum-resistant alternatives, global adoption remains insufficient, leaving sensitive data vulnerable to future compromise.
Accountability and Investment Gaps Hinder Progress
Uncertainty over who should own AI security responsibility further complicates response efforts. Twenty-nine percent of leaders assign accountability to the CIO, CTO, or technical teams; 26% to a dedicated AI leader; and only 17% to the CISO or cybersecurity function. Eleven percent report shared accountability across multiple functions.
Despite these gaps, financial commitment is increasing: 84% of security and finance leaders anticipate higher budgets, with 58% ranking AI as their top cybersecurity spending priority.
“Technology is moving incredibly fast, but the fundamentals of cybersecurity haven’t changed,” said Morgan Adamski, PwC’s cyber, data & technology risk leader. “You can invest heavily in AI and the latest security tools, but if you don’t have secure data, operational continuity, clear accountability and strong cyber hygiene underneath them, you’re building on a weak foundation. The goal isn’t to slow innovation down – it’s to make sure your organization is resilient enough to keep up with it.”
The report concludes that while the tools to strengthen defenses exist, widespread implementation is lacking. Surveys like this offer a snapshot of current perceptions, which may quickly become outdated given the rapid pace of AI and quantum advancements. Ultimately, the report urges organizations to focus on foundational security practices — encryption, access control, hygiene, and governance — as the essential base upon which advanced technologies like AI and quantum-resistant cryptography can be effectively deployed.
Source: SecurityWeek