As artificial intelligence evolves, its impact on cybersecurity and the workforce is profound and far-reaching. Predictive AI once allowed security teams to anticipate threats, and generative AI has taken creativity and automation to new levels. We are now on the threshold of agentic AI, an autonomous force capable of understanding tasks, making decisions, and taking action with minimal human intervention. This next phase of hyper-automation promises both unparalleled efficiency and a new set of challenges for organizations and society.
Richard Smithdirector of products and technology at SentinelOneis no stranger to this change. In a recent conversation, he described agentic AI as “the natural evolution of automation, in which systems not only execute workflows, but orchestrate and complete tasks autonomously while collaborating across fields.” As AI evolves, organizations must prepare for a world in which machines act not only as tools but also as collaborators in an increasingly complex cybersecurity landscape.
From hyper-automation to agentic AI: a new paradigm
For years, organizations have relied on playbooks, no-code workflows, and task orchestration to automate repetitive functions. These tools are effective but inherently inflexible, requiring manual updates and monitoring.
Smith explained: “Today’s hyper-automation works like a well-constructed playbook. This can trigger workflows, but ultimately it’s static. The next phase – agentic AI – is dynamic. Instead of following predefined instructions, these systems understand their domains, adapt in real time, and execute tasks autonomously.
Imagine an AI agent trained to respond to cybersecurity incidents. Instead of simply firing off canned responses, this agent could analyze data, collaborate with other AI agents, and adjust its approach depending on the situation. Such workflows are already being explored. For example, Smith explained that AWS security teams develop agentic systems that can act independently within their ecosystems.
The implications are staggering. Tasks that once required human oversight, such as resolving incidents, applying system patches, or scanning for threats, could soon be handled by AI agents operating seamlessly across networks.
Cybersecurity Threats in the Age of AI
While agentic AI is promising, it also amplifies existing cybersecurity challenges. Threat actors are increasingly leveraging AI to augment their attacks, creating a dangerous game of cat and mouse.
- AI-enhanced phishing: Generative AI tools like GPT-4 allow attackers to create highly convincing phishing emails, erasing traditional telltale signs like poor grammar and clumsy wording. Smith highlights the sophistication of these attacks: “We used to identify phishing by its obvious flaws. Now AI generates content so accurate that even savvy users can get caught up in it.
- Living Off the Land (LOTL): Threat actors exploit legitimate tools already embedded in enterprise environments. Smith describes this trend as a stealthy development: “Cybercriminals don’t necessarily deploy malware. Instead, they use what already exists, making detection much more difficult.
- State-sponsored intellectual property theft: Geopolitical tensions amplify cyber risks, particularly in countries like China. Smith warns of increasing attacks focused on intellectual property: “For many state actors, the goal is not just disruption. It’s about stealing intellectual property and gaining a long-term advantage.
To counter these threats, cybersecurity defenses must evolve alongside AI. SentinelOne integrates automatic triage and self-investigation into its platforms, enabling systems to autonomously identify and respond to threats. Additionally, Smith shared SentinelOne’s development of small language models (SLMs) for edge-based detection, thereby reducing false positives and improving performance. “SLMs allow us to bring intelligence closer to where threats occur, making responses faster and more accurate,” he said.
Labor and the automation debate
As AI advances, concerns about job losses are growing. For years, organizations have reassured their employees that AI will “enhance, not replace” human roles. Smith offered a more nuanced perspective: “AI will replace tasks, not people, at least in the short term. Human oversight remains essential because we do not yet fully understand AI behavior.
In cybersecurity, AI acts as a force multiplier, streamlining tedious tasks such as data analysis and incident documentation while allowing humans to focus on strategic decisions. This collaboration allows professionals to do more with less, increasing productivity without eliminating the need for human expertise.
However, Smith acknowledged the long-term challenges. As automation grows, society must grapple with economic changes, job losses, and the ethical implications of AI. Concepts like universal basic income may become necessary as machines take on routine roles.
“We are moving toward a future where AI will do the heavy lifting,” Smith said. “But we need humans for governance, creativity and innovation. The challenge is to ensure that society keeps pace with technology.
Preparing for the Age of Agentic AI
Organizations cannot afford to wait. To prepare for the rise of agentic AI and evolving cyber threats, businesses should prioritize:
- Strengthen identity security: With credential theft at the root of many breaches, organizations must adopt multi-factor authentication (MFA), access keys, and zero-trust architectures to secure access.
- Adopt AI-powered defenses: Leverage AI for threat detection, incident response, and automated remediation. Tools like SLMs can improve accuracy and reduce false positives.
- Balance automation and monitoring: While agentic AI can operate autonomously, human governance remains essential. Organizations must establish clear frameworks for monitoring AI-based systems.
- Invest in workforce training: Equip employees with the skills needed to work alongside AI. Rapid engineering, data analysis and AI governance will become essential skills.
A future of collaboration
The rise of agentic AI marks a transformative moment for cybersecurity and labor. As organizations move away from static workflows and adopt dynamic, autonomous systems, they gain the ability to respond to threats faster and more effectively than ever before. However, this evolution requires a strategic approach, one that balances automation with human oversight, strengthens defenses against AI-based attacks, and prepares for the societal changes that AI will bring.
Smith summed up the upcoming opportunity: “Agentic AI is not just the next phase of automation. This is an opportunity to redefine the way we work, secure our systems and innovate for the future. The question for businesses is not whether agentic AI will happen, but rather how prepared they are to adopt it.
I have a passion for technology and gadgets, with a particular focus on Microsoft and security, and want to help others understand how technology can affect or improve their lives. I also enjoy spending time with my wife, my 7 children, my 4 dogs, my 7 cats, a pot-bellied pig and a sulcata tortoise, and I like to think that I enjoy reading and playing golf even though I can’t find never time for one or the other. You can contact me directly at tony@xpective.net. To learn more, you can follow me on Threads, Facebook, Instagram and LinkedIn.
