According to a newly released report from Swimlane, a concerning 74% of cybersecurity decision-makers are aware of sensitive data being input into public AI models despite having established protocols in place. The report, “Reality Check: Is AI Living Up to Its Cybersecurity Promises?“, reveals that the rush to embrace AI, especially generative AI and large language models, has outpaced most organizations’ ability to keep their data safe and effectively enforce security protocols.
As AI becomes more integral to organizational operations, companies are grappling with both its benefits and the associated risks. To better understand this landscape, Swimlane surveyed 500 cybersecurity decision-makers in the United States and the United Kingdom to uncover how AI is influencing data security and governance, workforce strategies, and cybersecurity budgets.
“There’s no doubt that AI is reshaping cybersecurity as we know it, and its impact reaches far beyond the digital sphere,” said Cody Cornell, co-founder and chief strategy officer of Swimlane. “The fact that 74% of respondents view AI-generated misinformation as a significant threat to the U.S., particularly with the 2024 elections approaching, underscores the complex challenges ahead. While AI offers tremendous benefits in improving security and efficiency, it’s crucial that we approach its use responsibly, balancing innovation with the potential risks to both organizations and society.”
Key Takeaways
- Is AI Making It Impossible to Balance Innovation and Confidentiality? While 70% of organizations have specific protocols in place when it comes to what data is shared in a public LLM, 74% of respondents said they were aware of individuals at their organization inputting sensitive data into a public LLM.
- Who Should Govern AI? Only 28% of respondents believe the government should bear the primary responsibility for setting and enforcing guidelines. At the same time, almost half (46%) of respondents said the company that developed the AI should be held primarily responsible for the consequences when AI systems cause harm.
- AI Hype or Growth Engine? Seventy-six percent of respondents believe the current market is saturated with AI hype. This overload of AI-centric messaging is taking its toll, with 55% of respondents saying they are starting to feel fatigued by the constant focus on AI.
- Are AI Skills Essential? A significant 86% of organizations report that experience with AI and machine learning (ML) technologies significantly influences hiring decisions.
- Will AI Adoption Fuel Efficiency Gains and Increased Budgets? The majority of organizations (89%) report that the use of GenAI and LLMs has improved productivity and efficiency for their cybersecurity teams. This shift has led to a third (33%) of organizations planning to allocate more than 30% of their 2025 cybersecurity budgets to AI-powered or AI-enhanced solutions.
“Effective use of AI is no longer a luxury—it’s a necessity,” said Michael Lyborg, CISO at Swimlane. “By automating routine tasks and boosting threat detection, AI enables cybersecurity professionals to tackle more complex challenges head-on. Organizations that embrace AI strengthen their defenses and regain time for proactive threat hunting. As we navigate these turbulent waters, it’s vital that we implement AI thoughtfully to enhance security and uphold public trust.”
