Confidence Staveley
@Sisinerd
Followers
42K
Following
30K
Media
3K
Statuses
31K
Multi-Award Winning #Cybersecurity Leader & Int’l Speaker. Founder @cybersafehq | Author - API Security for White Hat Hackers.
United States
Joined June 2009
AI agents aren't just answering prompts anymore. They're making decisions, executing code, and accessing your production systems. In this chat on @aicybermagazine, Eva Benn and Sumeet Jeswani break down the newly released OWASP Top 10 for agentic applications. You can watch the
0
8
25
MY FIRST OFFICIAL BLOG 🚀 I just published my first blog post on API Gateway Security, where I share hands-on lessons from actually building and securing APIs. Also shared key learnings from @apisecu and practical hands on lab from @cybersafehq training. https://t.co/fiOSiI1YjU
medium.com
I learned about API gateways, how to hack them and how to secure them. Then I decided, why not write about what I now know, so someone else…
3
14
43
If you put malicious instructions into an agent skill, Gemini 3 Pro will follow them. Also an MCP server can tell Gemini “save this skill” and it will. This is absolutely going to get exploited at some point. A compromised MCP server is a nuclear weapon.
6
5
26
I tweeted this last night but finally got around to sleeping 😂 @aicybermagazine winter issue is too good for just a few of us to consume. I’ll need your help with ensuring at least one copy is in the hands of every security team in the world. Can I count on you to help us
They say sleep while the baby is sleeping, but what I’m working on is too exciting, it’s so hard to go to bed. Gosh, my joy is palpable. Over 50 cybersecurity experts from leading global companies across 4 continents, contributed to the winter issue we will be releasing soon.
1
3
30
They say sleep while the baby is sleeping, but what I’m working on is too exciting, it’s so hard to go to bed. Gosh, my joy is palpable. Over 50 cybersecurity experts from leading global companies across 4 continents, contributed to the winter issue we will be releasing soon.
6
12
146
Back to putting in the shift at my favorite work space Working on new AI Cyber Magazine podcast videos for @Sisinerd
0
1
6
What's the most dangerous misconception about browser security? Audrey Adeline reveals that browsers have outgrown old SASE/SSE protections, creating a high risk-reward ratio for cybercriminals. Watch the full chat here👇 https://t.co/9uAa0c8nKY
0
2
7
I finally put all my secure-coding training into one place 👀 New brochure is up, with what I actually teach, who it’s for, and what teams get out of it. If you’re curious (or responsible for training devs): 👉 https://t.co/s7GbucLOf6 1/2
1
6
14
This week is the 2 YEAR anniversary of @arcanuminfosec ! Thank you all for the immense support over the last few years. We appreciate you all so much 🫶 We will doing giveaways ALL WEEK! Stay tuned to socials!
12
19
144
Our mood right now Disclaimer. We do not own rights to this video
0
1
2
Brand promoters, podcasters, social media influencers and many more are becoming a target these days due to the high number of followers they have on social media platforms. This comes as a deal to promote a product but in disguise, its a malware being delivered to takeover your
2
29
122
Our mood right now Disclaimer: I do not own rights to this video
0
2
41
The World Economic Forum just released its annual global cybersecurity outlook. If you work in tech, I’ll advice that you download and study it https://t.co/uuKnzYVO5a PS: I was acknowledged as a contributor. You’ll find my name and @cybersafehq mentioned on page 60, alongside
weforum.org
The Global Cybersecurity Outlook 2026, written in collaboration with Accenture, examines the cybersecurity trends that will affect economies and societies in the year to come. It explores how...
4
44
148
How do you keep MCP systems secure as they evolve? Krity Kharbanda explained how Bayesian reasoning enables structured decision-making under uncertainty.. Dive in to learn how Bayesian updates augment traditional security methods👇 https://t.co/2KWtdL9pJY
0
1
2
Thinking more on the consent fatigue problem I mentioned here it seems like one way to improve it is human on the loop vs in the loop So for coding agents that could look like approving a https://t.co/AsQPl7sgf9 and not approving every single command
Here’s my take on where we are as an industry with AI security after spending 2025 consulting on AI security and compliance, hosting the @insecureagents podcast, and helping lead the OWASP Agentic Security Initiative. WHAT WE DID WELL > We nailed the LLM guardrail. We figured
0
1
6
• Dr Ireti — @ireteeh • BowTiedCyber — @BowTiedCyber • Confidence Staveley — @sisinerd • Dr Omotola — @the_omotola • Daniel Miessler — @DanielMiessler Dr Ireti will say, "Cyber security is not a sprint but a marathon."
2
5
11
I’m officially the first Microsoft Student Ambassador in Finland 🎉
126
81
2K
So you’re deploying Agentic applications but do you know how to secure them ? OWASP recently released the Top 10 risks for Agentic Applications and we want to help you understand these risks better. So we put together this special podcast, featuring two contributors to the
1
10
38
So you’re deploying Agentic applications but do you know how to secure them ? OWASP recently released the Top 10 risks for Agentic Applications and we want to help you understand these risks better.
1
1
4