Organizations today are increasingly recognizing the critical necessity of integrating artificial intelligence (AI) into their operations. While the promise of AI—particularly generative AI—can be revolutionary for enhancing internal productivity and customer-facing innovations, many organizations remain unaware of the cybersecurity risks associated with these technologies. The problem arises when these organizations, eager to harness AI’s potential, fail to ensure their deployments are secure, often leading to unanticipated vulnerabilities.
AI Adoption Outpaces Security Readiness
The enthusiasm for AI adoption is palpable. According to EY, a staggering 92% of technology leaders expect to ramp up their AI spending in 2025, marking a 10% rise compared to 2024. Agentic AI is emerging as a particularly transformative element, with 69% of leaders stating its necessity for staying competitive.
However, this rush to embrace AI often neglects crucial security considerations. The World Economic Forum (WEF) found that 66% of organizations anticipate significant impacts on cybersecurity due to AI within the next year, yet only 37% have mechanisms in place to assess AI security prior to deployment. Small businesses are particularly at risk, with 69% lacking safeguards for securing AI deployments, such as monitoring training data or maintaining an inventory of AI assets.
Furthermore, Accenture’s findings further underscore this gap: 77% of organizations lack foundational data and AI security practices, and merely 20% exhibit confidence in their ability to secure generative AI models. The implication is clear: organizations are diving headfirst into AI adoption without ensuring their data and systems are genuinely protected.
Acronis Cyber Protect Cloud integrates data protection, cybersecurity, and endpoint management.
Easily scale cyber protection services from a single platform – while efficiently running your MSP business.
Why Insecure AI Deployments Are Dangerous
Implementing AI without bolstering security can create significant compliance issues and serve as an inadvertent boon for cyber criminals. Insecure AI deployments empower cyberattackers who can exploit these technologies in multi-faceted ways:
- AI-driven phishing and fraud. The WEF highlights that 47% of organizations view AI-enabled cyberattacks as their top concern, a justified fear when considering that 42% faced social engineering attacks in the previous year.
- Model manipulation. Accenture’s research illustrates how AI worms like Morris II can implant malicious prompts within models, hijacking AI assistants to exfiltrate sensitive data or disseminate spam.
- Deepfake-enabled scams. Criminals are increasingly leveraging AI to generate convincing deepfake voices, images, and videos for fraudulent purposes. Notable incidents, like an attack that cloned the voice of Italy’s defense minister, exemplify the dangers of AI-driven deception, leading to financial losses for victims.
The landscape of cybercrime is evolving, with AI lowering the barrier for attackers, rendering scams faster, cheaper, and more challenging to detect.
Building Security Into AI from the Start
To harness the full benefits of AI responsibly, organizations must adopt a security-first mindset. This proactive approach entails embedding security at the outset rather than attempting to retrofit defenses after incidents or relying on a patchwork of tools. Here are practical strategies for achieving this:
- Embed security into AI development pipelines. Practices like secure coding, data encryption, and adversarial testing should be standard at each development stage.
- Continuously monitor and validate models. Ongoing testing of AI systems must be conducted to identify risks such as manipulation or data poisoning.
- Unify cyber resilience strategies. Security measures should be holistically integrated across endpoints, networks, cloud environments, and AI workloads to minimize vulnerabilities and ensure robust defenses.
Both WEF and Accenture stress that organizations equipped with strong cybersecurity capabilities and integrated strategies are best prepared for the AI advancements ahead. Accenture’s findings reveal that a mere 10% of businesses have reached what they term the “Reinvention-Ready Zone,” a space that combines mature cyber strategies with integrated monitoring, detection, and response frameworks. Firms in this category are 69% less likely to experience AI-driven cyberattacks compared to less prepared counterparts.
The Role of MSPs and Enterprises
With the surge in AI adoption, managed service providers (MSPs) find themselves at a critical juncture, facing both challenges and opportunities. As clients increasingly demand AI-enhanced solutions, they will also look to their MSPs to ensure security. According to the Acronis Cyberthreats Report H1 2025, AI-enabled attacks directed at MSPs have surged, with phishing attempts making up over half of all attacks in the first half of 2025.
This necessitates that MSPs provide integrated protection across cloud, endpoint, and AI environments, ensuring both their security and that of their clients.
For enterprises, the road ahead involves maintaining a balance between ambition and caution. While AI holds remarkable potential to enhance efficiency, creativity, and competitiveness, responsible deployment needs to be prioritized.
Organizations should elevate AI security to a board-level priority, establish clear governance frameworks, and ensure that cybersecurity teams are well-trained to tackle emerging AI-driven threats.
The Future of AI Deployments is Tied to Security
Generative AI is here to stay, likely becoming more entrenched in business operations. However, advancing without robust security measures is akin to constructing a skyscraper atop sand—a precarious foundation risks collapse.
By adopting integrated, proactive security measures and solutions, organizations can unlock the potential of AI without inadvertently broadening their exposure to evolving threats such as ransomware and fraud.
About TRU
The Acronis Threat Research Unit (TRU) comprises cybersecurity experts specializing in threat intelligence, AI, and risk management. This team investigates emerging threats, offers security insights, and provides IT teams with guidelines, incident response strategies, and educational workshops.
See the latest TRU research
Sponsored and written by Acronis.