Dr. Ed K. Kiyemba Edris, PhD

PhD▪CCISO▪️CISSP▪️CISM▪️FHEA▪️ Aspiring CXO▪ Cyber Security Leader▪ Technologist ▪ Educator ▪️ Research Scientist▪️ AI ▪️ cryptography▪️Science ▪Technology▪️PQC

[Dr. Ed K. Kiyemba Edris, PhD @edkedris]

I am always ready to serve, work, and collaborate as long as your goals/objectives align with mine. It is that simple.

[Dr. Ed K. Kiyemba Edris, PhD @edkedris]

• Vulnerabilities exploited 7 days before CVE disclosure. • LLM pipelines hit with 91k+ adversarial sessions in 3 months. • Malicious packages up 700%. • Breakout takes 27 sec; exfiltration starts in 4 mins. • Only 24% of enterprises have dedicated AI security teams.

[Dr. Ed K. Kiyemba Edris, PhD @edkedris]

1/ AI in OT/ICS: Threat actors use LLMs to accelerate recon on infrastructure.
2/ Trust as a Surface: Malicious Ads route users into fake AI chats that trick them into running CLI.
3/ Agentic Execution: Flaws in coding extensions show that as AI agents get local permissions.

[Dr. Ed K. Kiyemba Edris, PhD @edkedris]

AI is no longer a threat vector in theory. It's operational infrastructure for adversaries.
Three critical developments this week deserve serious attention from security leaders.
#Cybersecurity #Infosec #CISO #AI

[Dr. Ed K. Kiyemba Edris, PhD @edkedris]

Southern Africa ██████████████~14.5%
North Africa ████████████ ~12.5%
West Africa ██████████ ~10.5%
Central Africa ██████████~10.0%
East Africa ███████ ~7.5%

[Dr. Ed K. Kiyemba Edris, PhD @edkedris]

I decided to do a region. Generative AI Adoption by African Sub-Region

[Dr. Ed K. Kiyemba Edris, PhD @edkedris]

East and parts of Central Africa face the steepest barriers (connectivity, power, skills), reflected in lower clustered estimates.
Continent-wide, adoption remains below the Global South average (~14.1%), with DeepSeek helping boost access in restricted/low-cost areas.

[Dr. Ed K. Kiyemba Edris, PhD @edkedris]

Check out my latest article: Review of the UK Cyber Security and Resilience Bill (CSRB) 2025 and its Impact https://t.co/L4PW5B8SXJ via @LinkedIn

[Dr. Ed K. Kiyemba Edris, PhD @edkedris]

@elonmusk My account was suspended yesterday by your AI enabled system for posting security analysis on CAG vs RAG. Without even cross checking my profile or past historical posts. When you say critical thinking AI, I am just imaging kind of content that will be written. I might be wrong.

[Dr. Ed K. Kiyemba Edris, PhD @edkedris]

@twitter @grok who am I?

[Dr. Ed K. Kiyemba Edris, PhD @edkedris]

Real-time monitoring The future of AI security is not just hybrid, it’s hardened hybrid. hashtag#AI hashtag#Security hashtag#RAG hashtag#CAG hashtag#LLM hashtag#AISecurity hashtag#ZeroTrust hashtag#ContextEngineering hashtag#CyberAI hashtag#AIArchitecture

[Dr. Ed K. Kiyemba Edris, PhD @edkedris]

hashtag#Hybrid RAG–CAG Systems: Security & Risk Analysis (2025 Perspective)
As LLM architectures evolve, Hybrid RAG-CAG systems are becoming the new normal which combines:
- Static, trusted knowledge (CAG)
- Dynamic, untrusted retrieval (RAG)
- Cross-system orchestration

[Dr. Ed K. Kiyemba Edris, PhD @edkedris]

redacted, filtered, and sandboxed Orchestration: conflict detection and output Validation Hybrid RAG-CAG = Powerful, but risky. Hybrid and hardened means secure. Deploy only in Zero-Trust environments with: Signed & encrypted CAG Redacted RAG retrieval Verified sync pipelines

[Dr. Ed K. Kiyemba Edris, PhD @edkedris]

Kali GPT: the AI assistant transforming how you interact with Kali Linux. Tailored learning, enhanced productivity. https://t.co/0uFacdZfl3