There is a common pattern in modern-day phishing emails – most of them do not have attachments.
Why? Simply because, as an attacker, it is much easier to walk through the front door with stolen keys (or legitimate credentials) than to try to break in through security.
The use of stolen/compromised credentials was the top attack vector (22%) in reported breaches compared to exploitation of vulnerabilities. Traditional security is no longer reliable because it was built for a setup where identities and data remain “in the premises,” with the network being a fixed point.
However, the modern workplace is no longer like this, with hybrid teams working/collaborating across the globe beyond offices.
The attack surface has become borderless, with SaaS environments. Challenges have emerged, including rising risk of tech sprawl with complex setups, increasing vendors with direct access, and evolving attackers who know where to strike.
Through our blog, we will compare traditional security with modern zero-trust architecture and explain why zero-trust is effective. We will also evaluate where traditional can still work, along with the real costs and mitigation associated with zero trust.
What is traditional access control?
Traditional access control is a castle and moat security model, where trust is implicitly granted to users, devices, and processes within a defined network perimeter (akin to a moat) within an organization.
The perimeter comprises VLANs, firewalls, and network segmentation.
It was made with an assumption that everything and everyone inside the boundary is trusted. It is based on a single point of validation at the edge, where any user is authenticated first at the perimeter through VPN,firewall rules, or corporate credentials, and then allowed inside. Once inside, a user is, by design, trusted and can move freely from east to west without any large restrictions.
In the age of kingdoms and castles, the concept of building a moat around a castle was widely used as a measure to protect the castle. A moat is a deep, wide ditch surrounding a castle filled with water.
Traditional access control is similar. It was for a physical office setup where employees used to stay within the office.
The traditional model doesn’t work when your users are scattered across continents, your data is stored on the cloud, and you add new applications to your SaaS stack.
Here is where zero trust steps in.
What is Zero Trust? What are its core principles?
Zero Trust Architecture (ZTA) is a security model that requires verification of every access request regardless of its origin, eliminating implicit trust (hence the name “Zero Trust”). It requires continuous verification of access requests against identity, device health, and context before granting them.
NIST SP800-207 defines Zero Trust as a cybersecurity paradigm focused on resource protection and the premise that trust is never granted implicitly to assets or user accounts based solely on their physical or network location (i.e., local area networks versus the internet) or based on asset ownership (enterprise or personally owned) but must be evaluated continuously.
It specifies that authorization and authentication are discrete functions performed before establishing any session to an enterprise resource.
It is based on three key principles:
Never trust, always verify
- Zero trust requires real-time evaluation of each access request every time, throughout the session. It includes not just login but also user identity, device posture, location, and behavior.
Use least privileged access
- A user should only be given access for a specific task, in a specific session, for a specific resource, and for a specific reason. No blanket network permissions are allowed.
Assume breach
- The architecture was built with the assumption that the attacker has already infiltrated. For example, micro-segmentation helps limit blast radius, while logging monitoring and behavioral analytics help detect suspicious activities before they turn into an incident.
- Zero trust shifts focus from security controls based on the network perimeter to identities, requiring authentication and authorization.
Zero Trust vs Traditional Access Control: What’s the difference?
|
Aspect |
Traditional access control |
Zero Trust |
Winner |
|
Assumption of trust |
Trust is ASSUMED if a user is inside the network. |
Trust is NEVER ASSUMED. User/device/session is verified on every access request. |
Zero Trust |
|
Access type |
BROAD – users get network-level access after authentication. |
MINIMUM – users get the least privileged access needed to perform their task. |
Zero Trust |
|
Threat response |
REACTIVE – breaches are detected after lateral movement. |
PROACTIVE – blast radius is limited by micro-segmentation, and anomalies are flagged in real time due to continuous monitoring. |
Zero Trust |
|
Visibility |
LIMITED – blind spots are common with fewer insights on east-west traffic. |
BROADER – with insights across all the users, devices, apps, and data flows. |
Zero Trust |
|
Remote work fit |
Not built for hybrid/remote work. Relies on VPN bottlenecks. |
Effectively works from any location or device. |
Zero Trust |
|
Insider threat |
HIGH Exposure – insiders can leverage the access to move freely once inside the perimeter. |
LIMITED Exposure – least privilege and micro-segmentation can limit where insiders can move. |
Zero Trust |
|
Cost |
Lower upfront costs (licensing costs – firewall + VPN etc.) but higher long-term costs(from breach remediation). |
Higher initial investment in the ZTNA platform and IAM tools, but higher ROI over the years due to lowered breach risk. |
Zero Trust |
|
Compliance |
Only partially with requirements. Higher challenge meeting NIS2, CMMC, and DORA’s requirements. |
Natively aligns with SOC2, ISO27001, NIS2, HIPAA, DORA, and CMMC through logging (continuous monitoring) and built-in access controls. |
Zero Trust |
What changed in 2025-2026 that makes this comparison different?
Zero Trust has gone from being “nice to have” to becoming a necessity. Here is why:
Credential harvesting and abuse have risen
Attackers no longer need to exploit, as they can easily ‘buy’ harvested credentials and automate attacks using AI at scale. Perimeter defenses are failing as they cannot distinguish between genuine and malicious/stolen credentials, while zero trust ensures continuous verification of behavioral activity.
NIS2, DORA, and other compliances have changed expectations
The healthcare, transport, energy, digital infrastructure, financial, and other industries must adhere to the risk management and reporting obligations set by NIS2 and DORA that Zero Trust natively satisfies. It is also compliance-friendly with regulations, like SOC2, HIPAA, CMMC, and ISO27001.
Hybrid work has become more common
The hybrid operating model has become a new default, and traditional VPN centric architectures weren’t made for this. Zero trust solves this by verifying each access attempt regardless of the location. It enforces least privilege access, ensuring that only authenticated and authorized users have access.
Increased SaaS adoption has made the perimeter ineffective
When an organization has deployed multiple SaaS-based tools and collaborative environments, it has a perimeter-less environment on which traditional security might be ineffective because modern attackers use compromised identities. Zero trust enforces continuous validation of who or what is requesting access mapping to both NIST SP-800-207 and SaaS first-world security requirements.
Real-world case study – how zero trust helped where traditional access control failed?
The 2025 Kering incident is a real-world example of how traditional security has failed and why zero trust is critical for cloud setups.
Kering is the parent company of Gucci, Saint Laurent, and Balenciaga.
Through a single breach, attackers gained unauthorized access to systems associated with Kering. It exposed customer data from brands connected like Gucci, Balenciaga, and Saint Laurent.
The pattern was similar to a growing attack trend across industries – an attacker leveraging compromised credentials to gain unauthorized access, instead of making exploit attempts.
Without a single piece of evidence of any notable zero-day vulnerability exploited, the breach reflected a weakness common in traditional access control systems – security controls that trusted and offered access without any verification.
It is an example of how perimeter security fails when an attacker uses valid credentials. There was no verification, no detection of suspicious behavior, and no restriction on what anyone could access. An environment where access was granted solely on valid credentials, and where a compromised identity was treated as a legitimate identity.
The Kering data breach isn’t about technical complexity, but how attackers gained access without breaking in and just logging in.
If the organization had Zero Trust Architecture as defined by the NIST SP 800-207, the attacker would have failed in multiple layers. At the first layer, the MFA enforced would have ensured that stolen credentials couldn’t be used.
The continuous evaluation would have provided more context to the suspicious attempt, including geolocation, device posture, and behavioral insights, and allowed the blocking/flagging of suspicious attempts before they turn into a serious activity.
Lastly, the least privileged access controls would have ensured that large volumes of sensitive customer data cannot be accessed even by authenticated users without explicit authorization specific to their role.
With the Zero Trust model, high-risk actions are immediately blocked and flagged as suspicious, requiring additional verification from users before execution.
When traditional access control is still fine
Here is an uncommon opinion – zero trust isn’t always necessary. Traditional access control may be adequate in some cases.
For example, traditional access control may work if an organization has fewer than 20 employees, without any hybrid setup, employees working on-premises, without any remote access or SaaS tools, handling of sensitive information, or third-party integration. It is because the surface is small and the blast radius is containable.
One of the biggest mistakes is that organizations assume that these conditions describe their profile while having at least one SaaS tool deployed with compliance obligations requiring zero trust.
FAQs
What is Zero Trust Architecture?
Zero Trust Architecture is a security framework that eliminates all implicit trust from a network, where every access attempt is verified against the user’s identity, behavioral context, and device health. Its core principles include: a) never trust, always verify, b) Use least privileged access, c) Assume breach.
What is the difference between ZTA and ZTNA?
ZTA or Zero Trust Architecture is a security framework. ZTNA, on the other hand, is a specific technology that enforces the core principles of zero trust.
Is zero trust expensive?
The upfront cost of zero trust is more expensive than traditional access control (like VPN and firewall). For a mid-sized company with 200-500 users, it can cost around $80k to 120k per year. However, it can invert the costs associated with damage from breach and regulatory fines from regulations like NIS2, DORA, and HIPAA.
Can small businesses use Zero Trust?
Yes, but they would need assistance from companies like SharkStriker, who can help them implement zero trust tailored to their cybersecurity setup, size, and business at an SMB price point. They can start by implementing MFA across their organization and least privileged access before investing in micro-segmentation.
Does Zero Trust replace VPN?
It replaces the function of VPN (remote access) while securing the organization from lateral movement of the network. Most organizations migrating to zero trust run VPN in parallel with ZTNA until full migration.
How long does Zero Trust take to implement?
For enterprises, realistically, zero trust implementation takes around 12-18 months from scratch. However, meaningful security improvement (like implementing MFA) across an organization can take up to 30-60 days.
Is Zero Trust required for compliance
No regulation mandates zero trust. However, implementing zero trust can help adhere to multiple regulations, standards, and certifications (like HIPAA, SOC2, DORA, and ISO27001).