There is no getting away from the fact that passwords are still the cornerstone of modern cybersecurity practices. Despite decades of advice to users to always pick strong and unique passwords for each of their online accounts, Keeper Security found that only one-quarter of survey respondents actually do this. Many use repeat variations of the same password (34%) or still admit to using simple passwords to secure their online accounts (30%). Perhaps more worryingly, almost half (44%) of those who claimed all their passwords were well-managed also said they used repeated variations of them. One in five also admitted to knowing they’ve had at least one password involved in a data breach or available on the dark web.

At first glance, these results may come as a shock, especially to those in the cybersecurity industry who have been touting these simple best practices for years. However, when considering more than one in three people (35%) globally admit to feeling overwhelmed when it comes to taking action to improve their cybersecurity, and one in ten admit to neglecting password management altogether, the results are much less of a surprise.

Cybersecurity is a priority and cybersecurity solutions must also be. The threat landscape continues to expand as our lives shift from in-person banks, stores, and coffee shops to online banking, internet shopping, social networking, and everything in between. We have never been more dependent on our phones, computers, and connected devices, yet we are overconfident in our ability to protect them and willfully ignoring the actions we must take to do so. Perhaps we need more people to admit they’re as careless as a bull in a china shop, burying their heads in the sand like an ostrich or simply paralyzed with fear. Facing reality and coming to recognize what’s at stake, they can more confidently charge forward and take the necessary steps to protect their information, identities and online accounts.

• Why on-premises data centers still relevant, and how can they be more cost-effective
• How VergeOS integrates server virtualization, networking, and storage into a single unified code base, surpassing traditional HCI while supporting external storage.
• The three pillars of UCI are a unified code base, three-dimensional scaling, and virtual data centers.
• How VergeOS’s common sense, per server, licensing model saves you money.
• How VergeIO’s multi-layered approach to ransomware resiliency and data protection provides the fastest recovery options with the least data loss.

• Why VMware’s approach to SDN stalled adoption due to licensing fees, complexity, and hardware overhead.
• How VergeFabric, built into VergeOS, eliminates the need for dedicated SDN appliances or controller VMs.
• The advantages of VM-level segmentation and VDC-level security without VLAN sprawl or manual firewall rules.
• How VergeOS extends cost savings beyond the hypervisor by integrating compute, storage, and networking into a single platform.
• Why multi-site connectivity and disaster recovery are simplified with built-in SDN.

Forward Networks co-founder Nikhil Handigol explains how organisations can harness agentic AI in networking and security without jeopardising reliability.

What makes AI “agentic” – Unlike chatbots, agentic systems pursue goals, choose their own tools and act autonomously. Market researchers expect the segment to surge from US $5.1 billion in 2025 to more than US $47 billion by 2030, and Gartner predicts that one-third of enterprise software will embed such capabilities by 2028. 

Upside for networks – Agents can:
remediate newly disclosed vulnerabilities or block malicious traffic,
resolve connectivity issues and reroute flows around failures,
analyse patterns to predict and prevent impending outages.
 
These abilities promise dramatic efficiency gains for short-staffed IT teams and better user experience.        
Risks to manage – Because an agent can act without human oversight, a mis-trained model or bad data could break compliance, introduce downtime or even cause physical harm. “Trust but verify” must therefore guide every deployment.

Build the data bedrock – 
Autonomy is only as sound as the data it relies on. Enterprises need a complete, accurate record of every device, configuration and packet path. A network digital twin—a mathematically precise, continuously updated software replica of the live environment—provides that single source of truth.
 
Digital twins as guardrails
Pre-change simulation: Before any AI-driven or manual change, test it exhaustively in the twin to catch policy violations, compliance breaks or connectivity loss.

Continuous verification: Twin-based rules monitor live behaviour and alert operators to drift or emergent problems. With this safety layer, CIOs can accelerate agentic AI adoption while preserving control.

Bottom line – Agentic AI’s promise in networking is real, but benefits accrue only if robust data pipelines and digital-twin guardrails are in place. When those prerequisites are met, AI agents can handle routine operations and incident response, freeing humans for higher-value work and increasing organisational resilience.