Security assessments
Penetration testing for web applications, APIs and AI agents. Reproducible findings and practical remediation.
Explore assessmentsAbout OktsecCybersecurity.
Cybersecurity.
AI governance.
We help organizations find security weaknesses, govern AI and prepare their teams. Our work combines hands-on research with experience building and leading technology companies.
Working with businesses, governments and universities across Latin America and the Caribbean.
Partnership Catalyst Award.
Oktsec founder Gustavo Aragón received the award at the 24th Americas Competitiveness Exchange (ACE), presented by Leandro Rizzuto Jr., U.S. Permanent Representative to the Organization of American States (OAS).
The recognition reflects his contribution to regional collaboration through Oktsec, connecting cybersecurity, AI and professional training with businesses, public institutions and universities.
Read the award announcement
- Letters of intent signed
- 15
- Countries represented
- 10
Gustavo signed 15 letters of intent with people and organizations in 10 countries to explore collaboration in cybersecurity, AI and professional training.
ACE Barranquilla 2026September 28–October 2 · ColombiaAbout the program on the OAS website ↗
Program organizers & supporters
Test systems.
Govern AI.
Prepare people.
Work with Oktsec on a defined security need, a governance decision or a learning goal.
AI governance
Define who owns each agent, what it can access and which actions need review. Apply policy and retain evidence through Control and Cloud.
Explore governanceTraining
12 talks, workshops and technical labs, plus custom topics for businesses, governments and universities. Sessions delivered in Spanish.
Explore trainingRadar
Public security intelligence with source evidence, affected systems and response guidance. Focused on Latin America and the Caribbean.
Explore RadarSecurity for a world that runs on agents.
AI agents change what organizations need to secure. They can act across tools and systems, carry out tasks and pass work to other agents. Preparing for this requires clear limits on their authority, records people can verify and teams who understand when to intervene.
A world ready to work with AI agents.
Our vision is a world ready to work with AI agents, with the security controls and practical skills to use them responsibly. Across industries and countries, organizations need both systems they can govern and people prepared to supervise the work they delegate.
Make delegated authority enforceable.
Turn the permissions people grant into controls that hold when agents act. Our mission is to carry that authority across tools, systems and other agents: define the limits, enforce them at execution and preserve evidence people can verify.
Meet Gustavo Aragón
The experience behind Oktsec.
- Experience
- 20+ years in technology
- Leadership
- 14 years as CTO
- Security research
- 250+ findings reported to vendors
- International experience
- Argentina, Brazil, New Zealand and Hong Kong
Gustavo co-founded Brelo, a fintech startup in São Paulo, where he spent seven years as CTO developing and patenting low-level device-locking technology for loans secured by mobile phones. Today, he works full-time on Oktsec, drawing on that experience alongside his security research and open-source work.
Oktsec · Founder
Building Oktsec to help companies stay in control as AI agents gain access to more tools and systems. Oktsec checks actions against the team’s policies and keeps a verifiable record of each decision. Security assessments and dependency reviews help teams find weaknesses. Training helps business, government and university teams make informed decisions about cybersecurity and AI.
Banza/Adcap · CPTO
As Chief Product and Technology Officer, led product and technology in regulated financial services.
Brelo · Cofounder & CTO
Spent seven years building Brelo in São Paulo, Brazil, working with the team on patented device-locking technology for mobile collateral lending.
01
Recipient of the Partnership Catalyst Award at ACE Barranquilla 2026. The recognition and its significance.
02More than 250 security findings reported to vendors through security programs and coordinated disclosure, including Google, Microsoft, Stripe, Cloudflare, AWS and Mercury.
03AgentPay won first place at the Anthropic × Kaszek Latin America hackathon in 2026. The prototype applied deterministic checks to payments between agents.
04Inaugural Agentic AI Foundation Ambassador (Linux Foundation) and Claude Code Partner. Approved for OpenAI Daybreak Blue and Anthropic Cyber Verification defensive research access.
How we work with customers
What to expect from an engagement.
Agree the scope and permitted tests first. Your team reviews the findings and decides which controls to adopt.
Scoped Assessment
Start with one workflow. Explicit rules of engagement. Findings with executable evidence.
Customer controlled runtime
Enforcement stays in environments you own. No inbound requirement to run Oktsec Control.
Confirm and hold safety
Prove the path without escalating beyond the agreed scope. Keep a full trail after the run.
No secrets in web forms
Use forms to describe the workflow. Do not include secrets, credentials, customer data or source code.
Work you can examine.
Explore the research, customer example and open source projects behind oktsec.
Published research
Read the attack analysis, reproduction methods and sources in the research archive.
Customer use
See how Certuma combines agent assistance, policy checks and physician review.
Open source
Explore Aguara for dependency inspection and nanostack for structured development. How the projects fit together.