Krishna Gupta (@KrishnaGCEO) header
Krishna Gupta (@KrishnaGCEO) profile picture

Krishna Gupta

@KrishnaGCEO

Information Security begins with you. And I greet people with Namaste or Namaskara. An entrepreneur who provides Secure CEO as a Service™ to secure your risk.

 Bengaluru
krishnag.ceo
Joined June 2008
  • Tweets5.7K
  • Following183
  • Followers285
#Claude_Cowork: #Prompt_Injection → local-file exfiltration krishnag.ceo/blog/claude-cow…→-local-file-exfiltration/
Image from Krishna Gupta (@KrishnaGCEO)
3
Controlling Focus, Controlling Risk: How Executive Mindset Shapes #Cyber_Resilience Through #VAPT #AEV #Adversarial_Simulation #Adversarial_Exposure_Valiation krishnag.ceo/blog/controllin…
Controlling Focus, Controlling Risk: How Executive Mindset Shapes Cyber Resilience Through VAPT

Controlling Focus, Controlling Risk: How Executive Mindset Shapes Cyber Resilience Through VAPT

Most organisations believe cyber incidents come from technical vulnerability. In reality, it begins with misplaced executive focus.

krishnag.ceo
4
#AI_Management_Systems (#AIMS): A Practical Framework for Managing #Artificial_Intelligence #Super_Intelligence krishnag.ceo/blog/ai-managem…
AI Management Systems (AIMS): A Practical Framework for Managing Artificial Intelligence

AI Management Systems (AIMS): A Practical Framework for Managing Artificial Intelligence

An AI Management System is an organisational framework for implementing, maintaining & continually improving the governance of AI or SI.

krishnag.ceo
3
#Detection Opportunities that target #NAT krishnag.ceo/blog/detection-…
Detection Opportunities that target NAT

Detection Opportunities that target NAT

Connection tracking is central to modern NAT implementations and should be monitored continuously. Security teams establish baselines for NAT.

krishnag.ceo
3
#UDP Handling and Timeout Behaviour krishnag.ceo/blog/udp-handli…
UDP Handling and Timeout Behaviour

UDP Handling and Timeout Behaviour

Unlike TCP, the User Datagram Protocol (UDP) is a connectionless transport protocol. It does not establish sessions via a handshake.

krishnag.ceo
2
The Role of #Encryption in #Mitigating #Network_Layer_Attacks krishnag.ceo/blog/the-role-o…
The Role of Encryption in Mitigating Network-Layer Attacks

The Role of Encryption in Mitigating Network-Layer Attacks

End-to-end encryption remains one of the most effective for protecting the confidentiality & integrity, even when network is under attack.

krishnag.ceo
1
#Detection_Engineering and #Threat_Hunting for #NatJack krishnag.ceo/blog/detection-…
Detection Engineering and Threat Hunting for NatJack

Detection Engineering and Threat Hunting for NatJack

The disclosure of NatJack highlights the importance of monitoring not only traditional network traffic but also the behaviour of connections.

krishnag.ceo
5
How #NAT #Works #Internally krishnag.ceo/blog/how-nat-wo…
How NAT Works Internally

How NAT Works Internally

A NAT device is far more than an address translator. It acts as a stateful packet-processing engine that analyses traffic & creates records.

krishnag.ceo
6
#NatJack_Attacks: #Hijacking #TCP_Sessions and #Spoofing #DNS by Manipulating #NAT_Tables krishnag.ceo/blog/natjack-at…
Image from Krishna Gupta (@KrishnaGCEO)
7
From #Human_Capital to #AI_Infrastructure: Turning #Workforce Disruption into #Business_Ownership and #Strategic_Opportunity #AI #SI #Artificial_Intelligence #Super_Intelligence krishnag.ceo/blog/from-human…
From Human Capital to AI Infrastructure: Turning Workforce Disruption into Business Ownership and...

From Human Capital to AI Infrastructure: Turning Workforce Disruption into Business Ownership and...

An employee receives payments for performing an assigned role. An equity investor participates in the ownership of a business & the risks.

krishnag.ceo
2
#Zero_Click_Agent_Hijacking krishnag.ceo/blog/zero-click…
Zero-Click Agent Hijacking

Zero-Click Agent Hijacking

Zero-Click Agent Hijacking is an emerging cybersecurity —relevant to agentic AI that can autonomously read, call APIs, browse web, email.

krishnag.ceo
1
#APT #Repository #Trust_Error krishnag.ceo/blog/apt-reposi…
APT Repository Trust Error

APT Repository Trust Error

The APT NO_PUBKEY error is a security mechanism in Debian & Ubuntu that prevents a system from trusting software repositories unverified.

krishnag.ceo
16
#Open_Weight_Models: The Foundation of Private and Enterprise #AI #SI krishnag.ceo/blog/open-weigh…
Open Weight Models: The Future of Private, Enterprise AI

Open Weight Models: The Future of Private, Enterprise AI

An Open Weight Model is an artificial intelligence model whose trained parameters (known as weights) are publicly available for download.

krishnag.ceo
6
#Ollama: The Future of Private and On-Premises #Artificial_Intelligence #Super_Intelligence krishnag.ceo/blog/ollama-the…
Ollama: The Future of Private and On-Premises Artificial Intelligence

Ollama: The Future of Private and On-Premises Artificial Intelligence

Ollama is an open-source platform that enables you to download, manage, & run Large Language Models (LLMs) locally on your own PC/mac.

krishnag.ceo
7
#Fail2Ban Explained: Protecting #Linux_Servers from #Brute_Force_Attacks and Unauthorised Access krishnag.ceo/blog/fail2ban-e…
Image from Krishna Gupta (@KrishnaGCEO)
10
#LM_Studio vs #OpenClaw vs #Hermes Agent: #Local_LLMs and #Agentic_AI Are Not the Same Thing krishnag.ceo/blog/lm-studio-…
LM Studio vs OpenClaw vs Hermes Agent: Local LLMs and Agentic AI Are Not the Same Thing

LM Studio vs OpenClaw vs Hermes Agent: Local LLMs and Agentic AI Are Not the Same Thing

LM Studio helps you run the AI. OpenClaw & Hermes AI do things. The key distinction is between LLM inference & Agentic AI orchestration.

krishnag.ceo
21
#LM_Studio: A Complete Technical Guide to Running #Large_Language_Models #Locally #LLM #AI #SI krishnag.ceo/blog/lm-studio-…
LM Studio: A Complete Technical Guide to Running Large Language Models Locally

LM Studio: A Complete Technical Guide to Running Large Language Models Locally

LM Studio is a desktop application & local AI platform designed to make running Large Language Models on local hardware considerably easier.

krishnag.ceo
11
Add #security_headers in #Open_Lite_Speed (Web Admin GUI) & #command_line #Web_Server #HTTP3 #H3 krishnag.ceo/blog/add-securi…
Add security headers in Open Lite Speed (Web Admin GUI) & command line

Add security headers in Open Lite Speed (Web Admin GUI) & command line

In OpenLiteSpeed (OLS) documentation, the cleanest recommended place for security headers is the Virtual Host → Context configuration.

krishnag.ceo
16
#Security_Headers Explained: From Early #Web_Security Hardening to the #Generative_AI_Era #SI krishnag.ceo/blog/security-h…
Image from Krishna Gupta (@KrishnaGCEO)
1
#UFW - #Ubuntu #Firewall - #Allow_Listing/ #Safe_Listing #IP Addresses from #Cloudflare - #Content_Delivery_Network #CDN krishnag.ceo/blog/ufw-ubuntu…
UFW - Ubuntu Firewall - Allow Listing/Safe Listing IP Addresses from Cloudflare - Content Delivery...

UFW - Ubuntu Firewall - Allow Listing/Safe Listing IP Addresses from Cloudflare - Content Delivery...

Makes Cloudflare difficult to evade. CF WAF, rate limiting, bot controls & other edge protections only help when traffic passes through.

krishnag.ceo
9
Load more