White Paper

Data Security Directions Council Report

Data Security Directions Council Part 1 –

Redrawing the Trust Boundary in a Quantum and AI-powered world

2026 Data Security Directions Council Reports

AI has moved beyond experimentation. Organizations are rapidly shifting from pilots to operational deployment—embedding AI into products, customer interactions, software development, and internal operations.

But with AI comes a new security reality. AI systems now access sensitive information, generate recommendations that influence business decisions, write software, interact directly with customers, and operate with limited human supervision. At the same time, attackers are automating reconnaissance, scaling convincing phishing, speeding up vulnerability discovery, and deploying sophisticated attacks.

This challenges the trust boundary—the line between what we assume is trustworthy and what must be verified. This boundary used to separate verified actors from unverified ones, human judgment from machine output, and secure systems from exposed environments. Today, it’s being redrawn faster than established controls can adapt.

This report distills insights from practitioners responsible for securing, governing, and deploying AI across complex enterprise environments. Their insights are available in two parts:

Part 1: AI

Part 2: Quantum Computing

 

Download these Reports

AI

Humans are the bottleneck, and I don’t think we’ve fully realized this yet.”
Alain Michaud AI Strategy Lead Thales
The failure mode is not sudden collapse. It is a slow, invisible erosion of decision reliability.”
Aaron Kalvani Independent Global AI Strategist on Multilateral AI Projects United Nations
Every AI system should have clearly identifiable human and organizational accountability. Responsibility cannot be delegated to an algorithm.”
Anna Pouliou Chair & Data Protection Commissioner CERN
People confuse importance with urgency. AI for cyber is more important, but it’s less urgent. Ask everyone again in a year, and I expect their priorities will have shifted completely.”
Ofer Shaked Founder and CEO Deepblue Cybersecurity
I fear there are a lot of companies that haven’t realized that their greatest asset is their data. It is a cultural mind shift from thinking you produce widgets, to realizing you have a treasure trove.”
Steve Wright Data Protection Officer Financial Services Compensation Scheme
The last thing we want is to introduce governance functions that reduce the positive impact that AI can have.”
James Packer Cybersecurity Executive Ernst & Young

Our contributors disagreed on timelines, priorities, ownership models, and the pace of technological change. But they are aligned on what organizations should do next. Here are their recommendations:

  1. 1. Give AI Governance an Organizational Home
  2. 2. Build a Defensive Strategy That Assumes the Threat Will Outpace You
  3. 3. Resolve the Identity Problem for Non-Human Entities Before It Resolves Itself badly
  4. 4. Invest in Data Classification and Visibility as the Prerequisite for Everything Else
  5. 5. Move from Shadow AI Prohibition to Shadow AI Observation
  6. 6. Invest in Human Oversight Capability, Not Just Human Oversight Presence
  7. 7. Build Trust Through Behavior, Not Documentation

Data Security Directions Council Part 2 –

The Trust Boundary - Quantum Readiness and Data Security

2026 Data Security Directions Council Reports

Quantum readiness carries a strange asymmetry compared to most security priorities. The technology that will make current encryption obsolete does not yet exist at a meaningful scale. The data that will eventually break, however, is being collected today. Everything that follows is a response to that gap.

Priorities for Quantum Readiness

  1. 1. Complete a Cryptographic Inventory Before You Plan Anything Else
  2. 2. Make Crypto-Agility a Design Requirement
  3. 3. Prioritize Migration by Data Lifespan
  4. 4. Minimize Retained Data That Can’t Be Un-Harvested
  5. 5. Make Supplier Readiness Part of Your Own
  6. 6. Test Performance Before Migration
Download these Reports Today
 
If quantum-enhanced machine learning can break encryption faster than AI can automate the targeting, the reaction windows available to defenders will shrink dramatically.”
Ejona Preci Group CISO Lindal Group
You’ve got a door, it has a lock in it, you need to change the barrel of the lock because that type of key is now broken. It will cause companies to re-evaluate their data retention and transmission strategies. But it’s not rebuilding the door.”
Quentyn Taylor Senior Director – Product, Information Security and Global Incident Response Canon EMEA
Quantum computing alongside AI can be a cataclysmic paradigm shift. Very few organizations in the world can say that their systems can be protected also towards the use of quantum computing.”
Yiannis Kanellopoulos Founder and CEO code4thought
Partners Resources Blogs Sentinel Drivers