AI deserves our appreciation, but only if we're honest about what we're appreciating

AI Appreciation Day is a good excuse to acknowledge something true: AI has changed how organizations write code, analyze threats, and get work done at a pace that seemed unrealistic just a few years ago. That's worth recognizing.

Check Point's 2026 AI Security Report makes one point unavoidable: the same properties that make AI useful to defenders like speed, autonomy, the ability to operate with minimal supervision, are exactly what's made it so effective for attackers this past year.

From Assistant to Operator

A year ago, AI was described as a force multiplier, something that made existing attack techniques faster and cheaper to run. That description no longer fits what's being seen in the field. Recent research has documented intrusions where AI systems ran exploitation workflows with very little human direction, generating thousands of commands across dozens of sessions largely on their own.

In one case, a single developer used a commercial AI coding tool to build roughly 88,000 lines of functioning command-and-control malware in under a week, output that, on its face, looked like the work of a full team over several months. In a breach affecting Mexican government systems detailed in the report, one operator paired Claude Code and GPT-4.1 to compromise nine government agencies, turning about 1,000 typed instructions into more than 5,300 AI-executed commands. AI wasn't assisting that attack. It was running it.

Today, we see that the fine line between AI preventing attacks and acting as an attack agent is blurring. This is a call to action for organizations to see AI agents as a part of their attack surface.

The Attack Surface Nobody Budgeted For

The same capabilities that make AI agents genuinely useful, like reading documents, browsing the web, and connecting to other systems and tools, are also what make them exploitable. Hidden instructions buried in a web page or a calendar invite can hijack an agent's behavior. Configuration files that coding agents trust automatically can become delivery mechanisms for malware. A review of roughly 10,000 MCP servers found that about 40% carried security weaknesses of some kind, while published code packages continue to leak live credentials at a measurable, ongoing rate.

None of this is an argument against using AI. It's an argument for securing it with the same rigor applied to any system that touches sensitive data and critical infrastructure.

Enterprises Are Moving Faster Than Their Governance

The adoption numbers back this up. Organisations are now running an average of ten different AI applications a month, and in many cases without any formal approval process behind them. High-risk GenAI prompts, the kind that share sensitive corporate, personal, or regulated data with external AI services, have doubled over the past year, from 2% to 4% of all prompts. Between 87% and 93% of organizations had at least one high-risk GenAI interaction every single month.

That's not a call to pull back on AI adoption. It's a reminder that appreciation without governance is just exposure wearing better PR.

What Responsible AI Appreciation Actually Looks Like

If an organization wants to mark AI Appreciation Day the way a security-minded engineering team should, here's where to start:

• Treat AI as a live attacker when assessing your own defenses, not as background infrastructure you can set and forget.
• Assume your AI agents are targets, not just assistants. Prompt injection, poisoned configuration files, and runtime memory attacks aren't theoretical anymore; they're documented and repeatable.
• Govern AI usage the way you'd govern any other system that handles sensitive data. Employees are already sending sensitive information to external AI tools every day, policy or no policy.
• Match AI-speed threats with AI-speed defenses. A vulnerability can now go from disclosure to working exploit in hours. A human-speed security team, on its own, can't keep pace with that.

Building AI security capabilities, spanning AI Agent Security, AI Red Teaming, Workforce AI Security, and threat intelligence, comes down to a simple premise: appreciating AI and securing it are not competing priorities. They are the same job.