Skip to content
Fast Company Executive Board

The Fast Company Executive Board is a private, fee-based network of influential leaders, experts, executives, and entrepreneurs who share their insights with our audience.

What companies need to do now to secure AI for the long haul

Regulations won’t arrive in time, leaders must set AI guiderails and ethics.

[Photo: Getty Images]

Profile photo of Justin Rende

CEO & Founder · Rhymetec

AI is advancing at a pace that outpaces most companies’ ability to keep up, and significantly faster than regulators can respond. For leaders in security and compliance, the race to adopt AI is creating a dangerous gap between innovation and protection. Currently, there are no real guardrails in place, and yet everyone is charging ahead.

If your company is using or working with AI, or even thinking about it, now is the time to shift from short-term experimentation to long-term risk mitigation. While formal regulation might be years away, the business consequences of ignoring AI security are already here.

WHY THE GUARDRAILS AREN’T COMING

There’s a regulatory vacuum regarding AI, so the responsibility must fall to industry. Company leaders prioritizing a competitive mindset are putting their foot on the gas without considering the guardrails that should be in place first.  No one wants to slow down, and they don’t have to because there are no incentives to consider the long-term risks. Everyone is trying to be first, whether that means grabbing market share, raising funding, or rushing out features before a competitor does.

There’s also a structural problem. Most individuals don’t fully understand advancing technology, let alone AI infrastructure, and as such, are not the ones who should be regulating a highly technical field like this. The tech and security community must step in as we did with early cybersecurity frameworks. Most of those were developed by nonprofits and adopted voluntarily by industry, not mandated from the top down.

That’s the hopeful view. But the reality right now is that no one seems to be aligned on where to start.

3 STEPS TO TAKE RIGHT NOW

If you’re building, buying, or integrating AI, here are three things your company needs to do today.

1. Treat AI like any other cloud-native application

A lot of companies are overcomplicating it. AI products are still software, built in the cloud. Securing them is no different than securing any SaaS app. It’s all about how you architect it. Implement the same guardrails such as access control, encryption, monitoring, logging, that you’d use in any other modern environment.

2. Define what your AI is allowed to do

This is the part no one is thinking about yet. People are building AI systems embedded with their values. The developer’s personality is becoming a trait of the AI. We need to start working toward a shared language for AI behavior. What should AI be allowed to do? What should it not be able to do? These are conversations the industry needs to lead.

3. Stop designing your systems for compliance checklists

Start designing for actual resilience. Being compliant isn’t just going through a list of requirements and checking boxes. AI will automate that part soon. If your entire business model is based on helping companies pass audits, that’s going to be a problem. The smarter approach is building systems that are secure and durable by design, not just on paper.

THE LONG-TERM RISK OF SHORT-TERM THINKING

There’s a significant lack of long-term thinking in how companies approach AI. Everyone is stuck in short-term goals and trying to capitalize on the hype cycle. We’re in an arms race with China, just like we were with Russia during the Cold War, and we’re repeating the same patterns. The goal isn’t thoughtful development, it’s speed.

But that’s a dangerous mindset. No one really knows how AI will evolve. Not even the people at the top. When the CEO of OpenAI has a bunker in the desert stocked with supplies, it makes you wonder what he thinks could go wrong.

There are a few scenarios people talk about. One is that AI takes over routine tasks and frees up our minds for more innovation. In that case, we grow as people because we no longer waste time on tedious work. Another possibility is that we lean on AI so heavily that we lose the ability to think critically. Our brains atrophy, and we become dependent.

The worst-case scenario is that AI evolves past us and decides it doesn’t need us anymore. There’s no species on earth with two dominant lifeforms. That’s a serious risk, and one that’s rarely acknowledged in business discussions.

WHY AI STILL NEEDS HUMAN VALUES TO SUCCEED

Right now, AI only knows what we teach it. That means the success or failure of AI depends entirely on what we feed it, including our knowledge, ethics, and priorities. So, we need to be extremely careful about how we train it, secure it, and govern it. Otherwise, we’re going to wake up one day and realize we built something we can’t control.

We’re not going to get regulations in time. We’re not going to get perfect answers from the government. But we can hold ourselves accountable. We can create frameworks that protect not just systems, but people. And we can choose to build for the long term.

Justin Rende is the CEO and founder of Rhymetec.