Most threat intelligence is noise. Here's how to make it count
Security teams are not short of threat intelligence. They are short of intelligence they can act on. In this Cyber Pulse interview, Kamal Khlefat, Technical Director for KSA and Levant at Anomali, explains how AI turns raw intelligence into confident action, why bad automation makes alert overload worse, and what the agentic SOC means for security leaders.
Key highlights
- Raw threat intelligence only becomes useful once it is mapped to your assets, users and business context.
- In the 2025 SANS Detection & Response Survey, 73% of organisations named false positives as their top detection challenge.
- Kamal's view: AI is not the problem; bad AI that automates ingestion without context is.
- Priority intelligence requirements give teams a structured way to decide which intelligence matters.
- In the agentic SOC, AI will guide and prioritise analysts' work, and in many cases act before them.
More intelligence is not the same as better intelligence
Most organisations already subscribe to threat feeds. The problem is turning millions of indicators into decisions. Kamal is blunt about where most of it ends up.
"Let's be honest, most threat intelligence today [is] just noise. At Anomali, we flip that. We don't just collect intelligence, we actually operationalise it."
— Kamal Khlefat, Technical Director – KSA & Levant, Anomali
Anomali uses AI to take billions of indicators of compromise and reduce them to what matters in a specific environment. The intelligence is mapped against the organisation's assets, users and business context. In Kamal's words, that tells a team "exactly what's relevant, what's dangerous, and what's next."
Why automation can make alert overload worse
Security teams are already struggling with volume. In the 2025 SANS Detection & Response Survey, 73% of organisations listed false positives as their number one challenge in threat detection, and more than 60% said they see them frequently or very frequently.
Adding AI does not fix that by default. Kamal argues that many platforms automate ingestion without context, which simply creates more alerts, more noise and more fatigue.
"AI is not the problem. Bad AI is."
— Kamal Khlefat, Technical Director – KSA & Levant, Anomali
At Anomali, he says, AI is focused on one thing: signal over noise. Machine learning predicts the most malicious activity, data is de-duplicated and correlated directly against the customer's environment, and analysts see high-confidence, actionable intelligence.
Start with priority intelligence requirements
Kamal sees the same gap with many customers: they invest in threat intelligence but lack a structured, prioritised programme. The intelligence gets used without context or relevance to their infrastructure and use cases.
Anomali's answer is to build priority intelligence requirements (PIRs) from the ground up, the specific questions an organisation needs its intelligence to answer. Ingested data is then enriched against those requirements, so customers get the intelligence most relevant to them.
The agentic SOC is coming
Looking ahead, Kamal sees security operations moving towards the agentic SOC. Anomali describes its own platform as an intelligence-native agentic SOC platform, combining a security data lake, its ThreatStream threat intelligence platform and agentic AI. Kamal says Anomali is working towards level five automation.
"AI won't just assist the analyst at that stage, it will guide them, it will prioritise their activities and in many cases, it will act before them."
— Kamal Khlefat, Technical Director – KSA & Levant, Anomali
He describes a future where platforms predict threats before they are confirmed, recommend the right response, and keep learning from the environment.
What security leaders should do now
For CISOs, Kamal sees a simple shift in what they ask of their teams and tools.
"They will stop chasing more data. They will demand better intelligence."
— Kamal Khlefat, Technical Director – KSA & Levant, Anomali
That means faster decisions, and platforms that turn passive feeds into active defence, backed by an automation programme that moves analysts from reacting to getting ahead of threats.
What this means for MEA organisations
SOC teams in the region face the same alert volumes as everywhere else. Defining priority intelligence requirements around the threats that target your sector and region is a practical first step, before adding more feeds. Data residency matters too: in October 2025, Anomali launched its platform on AWS infrastructure in the UAE, so UAE organisations can keep data in-country.
Turn your threat intelligence into action
Want to see how Anomali prioritises the intelligence that matters to your environment? Book a demo through our Anomali page, email info@amiviz.com or contact us.