AI, Security, and the Future of Work: A Conversation with Valiant CIO Justin Penchina

There’s no shortage of opinions about AI right now. What’s harder to find is a conversation that’s technically grounded without being either breathlessly optimistic or reflexively skeptical.

On a recent episode of The Creative Stack, Georg sat down with Valiant CIO Justin Penchina to discuss what AI is actually useful for, where the real security concerns are emerging, and how small and mid-sized businesses can begin adopting these tools responsibly.

Justin has spent more than 16 years building and managing Valiant’s technology stack. When something genuinely complex needs to get done, he’s usually the person everyone calls. His perspective cuts through much of the noise surrounding AI and focuses on what matters most: practical business outcomes.

The most important distinction: AI as a tool vs. AI as the product

One of the most useful frameworks Justin shared is the distinction between using AI to do work and using AI to create the things people consume.

When AI helps organize data, summarize meetings, format reports, or generate a first draft, the value comes from completing a task more efficiently. Few people care whether a spreadsheet was manually reorganized or whether meeting notes were summarized by a human. The outcome is what matters.

The conversation changes when the content itself is the product. Creative work, thought leadership, and original writing derive value from human perspective, judgment, and experience. In those cases, AI can be an incredibly effective assistant, but replacing the human entirely may undermine the very thing that makes the content valuable.

Rather than asking whether businesses should use AI, the better question is: What role should AI play in this particular process?

Why AI feels different than previous technology shifts

If there’s one takeaway Georg wants every business owner to hear, it’s this: optionality creates leverage—and Justin compared today’s AI boom to the rise of cloud computing.

When cloud technologies first emerged, their transformative capabilities were largely available to enterprise organizations. Over time, SaaS platforms brought those benefits to businesses of every size.

AI is following a similar trajectory—but at a dramatically accelerated pace.

The difference is that AI arrived through an interface everyone immediately understood: a chat window. Tools like ChatGPT reached massive adoption faster than virtually any technology platform before them. What took cloud computing years to accomplish has happened in months.

The result is a technology shift that feels more disruptive, not because it’s fundamentally different, but because businesses have had far less time to adapt.

Where AI delivers real value today

Much of the public conversation around AI focuses on future possibilities. Justin’s focus is on practical applications that create value right now.

Meeting summaries are a perfect example. Instead of dividing attention between participating in a conversation and taking notes, AI can capture key takeaways, action items, and commitments after the fact.

The same applies to overcoming the blank-page problem. A rough collection of ideas can quickly become a structured first draft, allowing people to spend more time refining and improving their content rather than struggling to get started.

At Valiant, we’ve also used AI to streamline preparation of quarterly business reviews. What previously required exporting data from multiple systems, consolidating information, and manually formatting reports can now be completed in a fraction of the time.

Importantly, AI isn’t making strategic decisions. It’s handling repetitive, mechanical work that consumes valuable hours.

As Justin put it, AI’s greatest value often comes from shaving five or ten minutes off dozens of small tasks throughout the day. Individually, those gains seem minor. Collectively, they’re significant.

Understanding the rise of agentic AI

One of the most important developments in AI today is the emergence of “agentic” systems.

Unlike traditional AI tools that simply respond to prompts, agents perform actions on your behalf.

Justin recently built an agent that reviews his inbox every afternoon, identifies messages that still require a response, and provides a summary of outstanding items. Rather than waiting for instructions, the system proactively executes a workflow.

As AI becomes connected to business systems such as Microsoft 365, Google Workspace, Slack, project management platforms, and CRM tools, these agents will become increasingly capable.

For many small businesses, this may actually be an advantage. Unlike large enterprises struggling with fragmented data environments, many SMBs already operate within a relatively concentrated ecosystem of SaaS applications. That makes it easier to create useful AI-powered workflows without building massive infrastructure.

The opportunity is enormous—but so is the need for thoughtful security practices.

The security conversation most businesses aren’t having

Perhaps the most important part of the discussion centered on a technology many business leaders have never heard of: Model Context Protocol, or MCP.

MCP serves as a bridge between AI platforms and external applications. It allows tools like Claude, Copilot, or ChatGPT to communicate with services such as Slack, Microsoft 365, or project management platforms.

The challenge is that many MCP servers are not developed by the companies whose products they’re connecting to. Instead, they’re often created by independent developers and distributed as open-source projects.

That creates a potential trust gap.

You may trust the AI platform. You may trust the application you’re connecting to. But what about the software sitting in the middle?

Without proper vetting, organizations may expose sensitive information through tools they don’t fully understand.

This doesn’t mean businesses should avoid AI integrations. It means they should apply the same diligence, governance, and security scrutiny they’ve learned to apply to cloud platforms over the past decade.

The security ecosystem around AI is still maturing. Businesses that move thoughtfully will be in a much stronger position than those that move recklessly.

Three practical ways to get started

For organizations looking to adopt AI, Justin offered three straightforward recommendations.

1. Start with the tools you already own

If your organization uses Microsoft 365, begin with Copilot. If you’re invested in Google Workspace, start with Gemini.

These platforms operate within environments you already trust and manage. Your data remains inside existing security controls, making them a lower-risk entry point than introducing multiple third-party tools.

2. Provide context

AI outputs are only as useful as the instructions they’re given.

The more context you provide—who the audience is, what the goal is, what format you need, and what success looks like—the better the result.

Think of AI less as a search engine and more as a capable assistant. Clear direction produces dramatically better outcomes.

3. Treat every output as a draft

This may be the most important rule of all.

AI systems are remarkably capable, but they’re not infallible. They can confidently present inaccurate information, reference features that don’t exist, or make incorrect assumptions.

AI should accelerate your work—not replace your judgment. Review the output. Verify important facts. Edit before publishing.

The organizations seeing the greatest benefits from AI are not the ones blindly trusting it. They’re the ones using it to move faster while maintaining accountability.

The future isn’t replacement; it’s amplification

The question of whether AI will replace jobs inevitably came up.

Justin’s perspective was measured: some organizations may use AI as justification for reducing headcount, but the more common reality is likely to be increased productivity.

AI excels at repetitive and administrative work. Humans remain uniquely valuable for judgment, relationship-building, creativity, and strategic thinking.

In many cases, the organizations that attempted aggressive automation have discovered they still need people to guide, validate, and manage outcomes.

The businesses that thrive won’t be those that replace humans with AI. They’ll be the ones who figure out how to combine the strengths of both.

Final thoughts

AI is evolving quickly, but the fundamentals remain surprisingly simple.

Use it where it creates efficiency. Keep humans involved where judgment matters. Pay attention to security. Start with trusted tools. Verify the output.

For business leaders, the opportunity isn’t to chase every new AI trend. It’s to identify practical ways these technologies can improve operations, reduce friction, and help teams focus on higher-value work.

This conversation with Justin feels less like a conclusion and more like the beginning of a much larger discussion. As agentic AI matures and the surrounding security ecosystem evolves, we’ll continue sharing what we’re learning—and what businesses need to know to navigate the next phase responsibly.

The Creative Stack is produced by Valiant Technology, a managed IT services provider based in New York specializing in serving creative agencies and PR firms. Listen to episodes at podcast.thevaliantway.com and learn more at thevaliantway.com.