Artificial intelligence (AI) is rapidly reconstructing how governments think about identity. From defense to digital services, the questions remain the same. Who or what can be trusted to access critical systems, and how should their identities be managed in a world that is increasingly dominated by non-human actors?
This was the central theme of the Think Digital Identity and Cybersecurity for Government panel discussion. It featured former Identity Strategy Officers Howard Tweedie, Jonathan Neal, Jonathan Neal, Saviynt Field CTO, Ian Norton, and Digital Identity Advisor from Digital Identity Advisor.
From platform to decision
Tweedie opened by reflecting on lessons learned from military campaigns in Libya, Syria, Ukraine and other places. Civilian networks and data currently support military operations. “Platforms, tanks, trains, and very expensive elements are motivated by the way they integrate them into sensations, decisions and actions. We are moving from platform-centric to decision-centric paradigms.”
He argued that the change has only accelerated since 2021 due to the explosion of social media data and the integration of AI, machine learning and automation of robotic processes.
Both sides of the coin
In Neil's case, AI and identity must be seen in two dimensions: AI identity and AI identity.
First, he said, “AI has a major impact on how efficiency can be improved throughout the lifecycle management process.” “Reduce manual and common tasks, increase accuracy when making access decisions, and maintain compliance at scale.”
However, the difficult task is the opposite. This is a way to establish the identity of the AI agent itself. “When you're dealing with humans, you have HR. You do background checks, security clearance, criminal history. But when it comes to AI agents and machine identities, there is no HR. But these entities have access to critical systems and data.
Neil warned that this is no longer a future issue. “As already more than 60% of all internet traffic is from machines to machines. In businesses, the ratio of human-to-human identity to human identity is about 45:1, and has risen. Anyone who thinks this is the problem tomorrow is wrong.”
Old questions, new tools
Norton urged his colleagues to get through the hype. “The fundamental question hasn't changed. Who are you, can I trust you, and can I give you access?” he said. ” what It hasn't changed, but how I have it. Identity is used to mean appearing on a passport and catching your eye. Today, it means digital evidence, verifiable credentials, and a decentralized identity. ”
If you like this content…
Norton argued that the real breakthrough was “the identity of things.” “We no longer control people. We need to have the systems we discussed, the AI agents talking to the AI agents and hundreds of thousands of agents and manage all of those identities.
He explained the risks in his warning stories. The productivity app that uses AI to manage email worked well up to version 16, which told me to “open all emails.” “AI just did that. They didn't say checks, balances, or humans. That's the problem.
Zero Trust, Ethical Questions
For Tweedie, the answer lies in adopting zero trust, but it also expands its pool of expertise. “It's not just the technicians anymore. Maybe the social scientists on your team need to say.
In defense, it extends to whether AI can be trusted to make targeting decisions. “Do you automatically prosecute targets based on AI information or do humans need to be in the loop? They are no longer theoretical questions. They are alive.”
Clarity, the result, guardrail
When asked how the government should respond, Norton urged pragmatism. “Make sure you really clarify your use case. What outcome do you want to achieve? Then install the right guardrails – standards, policies, technical controls – you can innovate safely. But AI is still Wild West.
Neil agreed, pointing out the need for observability. “Even with the best intentions, exceptions arise. It requires continuous discovery and validation of identity. And the only way to manage AI at scale is to use the AI itself.”
New Identity Frontier
All three panelists agreed that the government needs to move quickly to adapt. “There are no more borders around these things,” Norton warned. “It's not just the government, it's not just the department. It's cut from one another, it's global. We all use the same technology, whether it's Spotify or for defense. So we're all facing the same risks.”
As Tweedie summed up, “AI gives us flexibility and speed. We can manipulate it in days rather than months. But the outcome is profound. Leaders have to ask. Is it really ethical?
