Advances in artificial intelligence (AI), coupled with increased accessibility and sophistication of AI-based impersonation attacks, have elevated AI to the top data security concern.

According to a new survey conducted by S&P Global Market Intelligence for Thales, AI-fueled attacks are now a prominent threat, with 59% of the 3,120 global security and IT experts polled reporting witnessing deepfake attacks, underscoring the rapid proliferation of AI-based social engineering.

The cost and impact of these technologies is becoming tangible. 48% of respondents said that they have experienced reputational damage as a result of AI-generated misinformation. This reveals how such content can inflict harm on personal and professional reputations, leading to significant credibility loss, eroded trust, and potential financial consequences.

Infrastructure supporting AI applications and data are also emerging as a prime target for attackers. 70% of respondents cited the rate of change as the top AI risk, suggesting that the speed of evolution within the AI ecosystem is a primary concern. Furthermore, 61% reported that their AI applications are being actively targeted, with sensitive data being the leading target.

Against this backdrop, spending on AI security is accelerating. According to a recent 451 Research Voice of the Enterprise study by S&P Global Market Intelligence, AI security spending now ranks as the second-highest priority, trailing only cloud security, among 17 listed areas. One in four respondents ranked AI security among their top three spending categories.

Furthermore, 30% of organizations now have a dedicated budget for AI security, up 10 points. However, more than half (53%) still fund AI security using their existing security budgets. This indicates that for most organizations, AI security is not yet treated as a distinct priority, suggesting that tools and strategies may still be lacking the specialized capabilities needed to detect, mitigate, and respond to unique AI risks.

Adoption of AI surges

Rising security concerns coincide with the massive surge in AI adoption within the business community. According to the 451 Research Voice of the Enterprise study, more than a third of enterprises (34%) reported that embedded agents are already in use, and a majority (73%) said they expect to use them within 12 months.

These findings align with Deloitte’s 2026 State of AI report, which is based on a survey of 3,235 director-level to C-suite-level respondents globally in Q3 2025. The study found that workforce access to AI expanded by 50% in 2025, growing from under 40% a year prior to under 60% of workers with sanctioned access to AI tools. Notably, 11% of leading companies now provide workers with near-universal access to sanctioned AI tools.

The study also found that AI adoption is rapidly shifting from experimental pilots to production. 25% of respondents indicated that their organization have moved 40% or more or their AI experiments into live products so far. 54% expect to reach that level in the next three to six months, demonstrating the pathway to value is clear and achievable.

The survey reveals that AI deployments are already yielding benefits. 25% of the leaders reported that AI is having a transformative effect on their companies, more than double from 12% a year ago.

In particular, 66% cited widespread gains in efficiency and productivity. However, benefits in other areas are taking longer to achieve. For example, revenue growth remains largely an aspiration, with 74% of organizations hoping to grow revenue through their AI initiatives in the future compared to just 20% having actually achieve this. Similarly, while 65% of respondents aim to reduce costs, only 40% are seeing these benefits realized.

The rising threat of quantum computing

In addition to AI risks, the Thales and S&P Global Market Intelligence study emphasizes the growing threat landscape surrounding quantum technology, identified by 61% of respondents as a top issue.

When asked about specific quantum risks, 61% of respondents cited future decryption of existing data, known as harvest now, decrypt later (HNDL), as the top concern. The second-highest concern, future encryption compromise, was the top concern in the year-ago survey.

Organizations are beginning to prepare of these risks. The survey examined organizations’ progress in mitigating quantum risks, as reflected in planned security measures for the next 18-24 months, and found a slight uptick in the proportion of organizations prototyping and evaluating post-quantum cryptographic (PQC) algorithms (59%).

Quantum technology harnesses quantum mechanical phenomena like superposition and entanglement to develop advanced devices and systems. Quantum computing is one of the most prominent applications of quantum technology, promising exponential speedups for complex problems, such as molecular simulation, cryptographic analysis, and risk model optimization, that classical computers cannot efficiently solve.

However, quantum computing also introduces risks, including the potential to break widely used cryptography methods such as RSA and ECC, which have protected sensitive data for now decades.

Despite these risks, organization are actively working on quantum computing projects. Almost a third (32%) of the respondents of the Thales and S&P Global Market Intelligence study reported having either already identified quantum projects or started experimenting with quantum computing.

Respondents anticipate the most significant impacts of quantum computing as improvements in machine learning, such as enhanced classification (57%), as well as advanced simulation (54%) and concurrent data processing (49%).

Featured image: Edited by Fintech News Switzerland, based on image by yanalya via Freepik