Pathway: Secure & Scale · Domain: Security & Resilience · Being developed by FajrFiber
Security & Resilience · Network Security & Threat Defense
AI-assisted Network Threat Defense
IN DEVELOPMENT
FajrFiber is developing this capability. It is not yet available as a service.
Layered protection being developed for enterprise networks: Secure DNS, intrusion detection and prevention, real-time threat analysis and behavioral analytics, designed so that AI assists analysis and response and policy governs what happens on your network.
Security decisions are network decisions.
AI-assisted Network Threat Defense is a managed enterprise network-security capability that FajrFiber is developing, extending its network engineering and operational responsibility into threat detection, evidence and governed response.
FajrFiber already designs, connects and operates business networks. A response takes effect on that same network and must be scoped, accountable and reversible where applicable.
Today
- Secure network architecture
- Network-security services within agreed scope
- A network security layer operated within FajrFiber’s own infrastructure
- NOC-backed network operations and engineering escalation
IN DEVELOPMENT
- Secure DNS
- Intrusion detection and prevention
- Real-time threat analysis and behavioral analytics
- AI-assisted analysis
- Governed response and evidence for your team
Layered, not single-point.
No single control sees everything. The capability is being designed as layers that each contribute evidence, with response kept separate from detection.
- Detection
- Is being designed to identify known attack patterns and unusual behavior on the network.
- Analysis
- Is being designed to correlate evidence across layers, explain it and prioritize what matters.
- Response
- Is being designed to act on the network only through policy, scope and approval where required.
Being designed to observe network evidence, not to forward traffic.
Separate, governed authority over enforcement.
The capability is being designed with three separate planes: traffic/data for connections, security telemetry for observing evidence, and response/control for enforcement through policy and approval where required. Analytics, AI and workflow automation do not receive unrestricted production authority.
Edge protection and intrusion prevention
Being designed to enforce agreed security policy where the network meets the internet, including intrusion prevention for known threats.
Secure DNS
Being developed as a first-class control that declines to resolve domains in agreed security categories before a connection is made.
Intrusion and threat detection
Being designed to detect known attack patterns and suspicious activity on the network.
Behavioral analytics
Being designed to highlight meaningful changes from the network’s usual behavior.
Real-time threat analysis
Being designed to bring security events from each layer together and analyse them as they arrive.
AI-assisted analysis
Being designed to correlate evidence, prioritize attention and recommend responses.
Governed response
Being designed so that every action passes tenant and action scope, confidence, rate limits, approval and audit controls, with expiry for temporary containment and rollback where applicable. Recommendation, not command.
Evidence and visibility
Being developed to show your team what happened, what was done and why.
A component in development
Secure DNS: a first-class control.
Secure DNS is being developed as a first-class protective control: resolving an organization’s queries and declining to resolve domains in agreed security categories, such as known malware, phishing or command-and-control infrastructure.
- Separate policies. Security categories and organizational acceptable-use rules remain distinct.
- Governed exceptions. Allow and block requests are being designed to be reviewed, recorded and, where appropriate, time-limited.
- Network-level evidence. Visibility is being designed around sites and network segments. Personal browsing histories are not a default feature.
Watching the network, within honest limits.
The capability is being designed so that network threat detection looks for known patterns and highlights unusual connections, destinations or volumes for investigation. It does not claim visibility into encrypted payloads. An anomaly is a reason to investigate, not proof of a threat.
No detection approach finds every threat. Relevant visibility limitations must be disclosed; missing telemetry only lowers or caps confidence when it matters to the detection being evaluated.
Governed response by design
AI assists.
Policy governs.
AI is being designed to assist detection, prioritization and response decisions. It is not the root of trust. The capability is being designed so that FajrFiber security engineers and authorized customer roles govern high-impact actions where required.
- 01
A response is recommended
Recommendations may originate from AI-assisted analysis, deterministic rules or FajrFiber security engineers. Actual provenance is retained.
- 02
Policy evaluates
Confidence, affected scope, impact modifiers, action class and customer policy determine authorization.
- 03
A person approves where required
High-impact actions require human approval.
- 04
The system executes
Only the authorized action, within its approved scope and applicable time limit.
- 05
The outcome is verified
Checked, recorded and reversible where applicable.
Confidence is not permission. Affected scope and business impact are different. An action on one critical device is still one-device scope; criticality increases its authorization requirements.
Temporary containment expires. Persistent policy changes require explicit approval, version history, an audit record and a rollback or reversal path; they need not have a temporary-action expiry.
Accountable people, defined before availability. The capability is being designed so that FajrFiber security engineers and your authorized roles govern high-impact actions. The capability is being developed alongside FajrFiber’s NOC-backed network operations. Roles, hours and response terms will be defined in service terms before it is offered.
Four questions, answered with evidence.
Customer security visibility is being developed alongside the capability. It is being designed to keep detected events, proposed actions and executed mitigations clearly distinct.
- 01
What happened?
The detection, its severity and what it affected.
- 02
What did FajrFiber do?
Recommendations, decisions and actions, with their real origin and timing.
- 03
What is the evidence?
The signals behind each conclusion and relevant visibility limitations.
- 04
What should you do next?
Specific next steps for your team.
Tenant isolation and least privilege by design.
The platform is being designed for strict tenant isolation, role-based access, MFA / SSO readiness and audit history.
- Low-sensitivity defaults. Read-only access defaults to network/security visibility. Detailed device or activity context requires an explicit permission or role grant.
- Privacy by design. Employee-level activity is not a generic surveillance feature. Identity-linked detail requires organizational policy, a lawful purpose and an authorized role.
- Accountable access. Policy changes, approvals, actions and access to sensitive detail are being designed to be auditable.
Clear about the limits.
- AI does not have unrestricted control of the network.
- No guarantee of detecting or preventing every threat, including unknown threats.
- No claim of complete encrypted-payload visibility.
- Employee surveillance is not a generic feature.
- FajrFiber is developing this capability. It is not yet available as a service.
How it is being developed
Built in stages, proven before it is offered.
IN DEVELOPMENT
Controlled enterprise pilots are planned before production availability. The sequence below describes the development path; it does not mark completed stages.
- Lab telemetry
Build and test the evidence pipeline.
- Secure DNS
Develop a first-class protective control.
- Detection
Network detection and centralised telemetry.
- Shadow automation
Review recommendations without executing them.
- Governed response
Bounded actions with appropriate approval.
- Enterprise pilot
Controlled pilots before production.
- Production
Service after readiness approval.
Network security, today.
While this capability is developed, FajrFiber designs secure network architecture and provides network-security services within agreed scope. Start with the network you have.