Back to home
AI

Secure and Responsible AI

Adopt AI boldly, then close the gap between how fast you adopt it and how fast you can control it.

I lead Accenture's Secure, Responsible AI and Data Protection practice across the Americas: the United States, Canada, and Latin America. The work runs on three fronts. Security for AI puts the governance, guardrails, and evaluation in place so the business can use GenAI and agents without inheriting unmanaged risk. AI for Security rebuilds the security operation around AI because the adversary already has. Data Protection shrinks the blast radius when something gets through anyway.

What I bring to it is the operator's view. I have sat in the CIO chair during a technology shift and know that the danger is rarely the technology itself. It is the gap between adoption and control. The interactive labs on this site exist so that executives can feel that gap rather than read about it: break a hardened assistant, run an incident from the executive seat, red-team your own AI policy.

Three fronts

Security for AI

Adopt AI boldly, with a foundation built to contain its risks.

GenAI and frontier models are reshaping how the business operates, but they also open a new attack surface: prompt injection, model and data poisoning, sensitive-data leakage, and agents that act with real privileges. I help organizations put the foundation, guardrails, and controls in place, including AI governance, model and pipeline security, runtime guardrails, and continuous evaluation, so the enterprise can capture AI's upside without inheriting unmanaged risk.

AI for Security

Attackers already use AI. Your defense has to answer in kind.

AI has lowered the barrier to entry for attackers and collapsed the time from intent to impact. Defending at that speed means rebuilding security operations around AI rather than bolting it on. I help organizations redefine the methodology and processes so AI accelerates detection, triage, investigation, and response, through agentic SOC workflows, automated threat analysis, and human-in-the-loop decisioning that keeps analysts focused on what matters.

Data Protection

You can't always prevent the breach. You can shrink the blast radius.

Every AI and digital initiative ultimately runs on data, which makes data both the prize and the liability. I help organizations minimize exposure across the data lifecycle with the full set of data protection disciplines: data security and classification, encryption, tokenization, and minimization. The goal is simple: reduce how much sensitive data is exposed, and contain the damage if an attacker ever gets through.

What I have done

Cinch: a production collaboration platform built on evidence, not demos

Independent · 2025 to present

A multi-tenant Slack alternative in production at cinchme.app, with tenant isolation proven in the database, a browser suite run against the production image, and a release gate that refuses to promote traffic without evidence.

2,566
Unit and integration tests
92
Browser scenarios against the production image

AI-assisted product development, 90% faster to market

North Highland · 2023 to 2026

A product development methodology built around AI-assisted engineering that reduced cost and time to market by more than 90%.

90%+
Reduction in cost and time to market
[verify]
Number of products or accelerators delivered

Proof

3
Countries and regions: US, Canada, and LATAM
3
Practice fronts: Security for AI, AI for Security, Data Protection
6
Interactive AI security labs on this site
90 sec
To locate your organization on the AI Control Gap model

Writing

  • Securing the Enterprise Rush to GenAI

    Adoption is outrunning governance by a wide margin. The defining risk of this era is not an exotic zero-day. It is the debt between how fast we deploy AI and how slowly we learn to control it.

The framework

The AI Control Gap

The danger isn't the AI. It's the gap between how fast you adopt it and how fast you can control it. See where your organization sits in 90 seconds.

Find your gap

The lab

Tools and frameworks

  • NIST AI RMF
  • ISO 42001
  • EU AI Act
  • OWASP LLM and agentic Top 10
  • AI governance and guardrails
  • Model and pipeline security
  • AI-augmented SOC design
  • Encryption, tokenization, data minimization
  • AI-assisted development