I was recently asked to speak at an event but was unable to accept due to a prior speaking commitment on the same day. This, in and of itself, is not unusual – it happens fairly often. What happened about a week later, however, gave me some interesting food for thought.
Someone that was available and had agreed to speak at the event in my place approached me to ask for content for the talk. Now, I am always happy to help someone work through what type of content might be interesting and appealing to the audience, as well as to point them in the right direction regarding developing that content. The manner in which this person approached me was not that, unfortunately.
In the moment, I tried to guide the individual in a helpful way. Only later, after analyzing the interaction a bit more, did I realize something. Essentially, the person approached me as people approach AI. How so? Rather than ask for guidance, tips, lessons learned from experience, dos and don’ts, etc., the person basically sent me a prompt. Something that when I thought about it later sounded a lot like “Tell me the most interesting points on topic X and provide me a presentation on that topic.”
This experience got me thinking. Are there ways in which AI could actually be harming us as a security industry? Of course, I know that there are many use cases for which AI helps increase productivity, efficiency, and has other benefits. I am also aware that there are many instances in which attackers are using AI to increase the speed and accuracy with which they attack organizations. What I began asking myself was something fundamentally different – could AI actually be harming the security industry?
The conclusion I came to was yes, it is. While this may sound like a radical stance in 2026, allow me to illustrate through several examples:
- Creativity: While AI can rapidly curate and search through a tremendous amount of data, it is no substitute for human creativity. Most of us recognize the value of creativity in art, music, writing, and other areas. But how many of us appreciate creativity in the security profession? In other words, the creative ways in which security professionals have discovered new things, approached challenges, solved problems, observed irregularities, and noticed oddities? When we use AI for certain security tasks, we lose out on human creativity and all of the benefits that creativity brings, much to our detriment. In my opinion, certain tasks just can’t be taken from a human and given to AI because of this.
- Humor: Humor can be a great way to make a point, diffuse tension, break bad news, and a whole host of other things. Because of this, it is a soft skill that has helped security professionals operate more effectively for decades. When we use AI to create talking points, write for us, create presentations for us, or otherwise, we give up any and all chance for witty humor. I believe this does us a disservice as a profession. Security, like many serious topics, is a medicine best served with a dose of humor to help it go down easier.
- Emotion: AI can mock, mimic, or play to certain human emotions, but it does not feel those emotions. That makes AI feel insincere by nature. Humans are often motivated by sincerity. Many of us are inclined to try harder for people who approach us with genuineness, even if we don’t necessarily want to do or agree with what they are asking of us. The security team that overrotates on AI loses sincerity – arguably one of the most important tools it has.
- Messaging: Recently, while attending a large security conference, I heard quite a few people note that most of the companies on the expo floor were saying all the same things. While this is partly driven by marketing departments looking to use buzzwords that attract attention and drive engagement, there is another aspect here as far as I can tell. It seems that far too many marketing teams are using AI to generate content. Not surprisingly, the messaging that results is unremarkable, non-unique, and woefully unprovoking. We as an industry need messaging that inspires thought, sparks conversation, and remains with people. That means moving away from using AI to help us message.
- Communication: Unfortunately, I’ve noticed a trend towards language becoming less precise and less clear in recent years. There are various forces that I can think of that are contributing to this decline, but AI is definitely one of them. It’s not hard to identify when something has been written by AI, in part or in whole. Security is a profession that is highly dependent on timely and accurate communication. Sadly for us, AI is weakening our communication skills, which is damaging our profession as a whole.
- Critical thinking: In the security field, as in many fields, there are just some situations that no one has encountered before, and for which there is no pre-existing answer. These situations require critical thinking ability. Unfortunately, with so many people in our profession looking to AI to perform analysis for them, offer them suggestions, and organize their thoughts, this important skill is fading. The security profession with an impaired critical thinking capability is not the security profession most of us grew up in.
While the picture may seem grim, all hope is not lost. We can establish firm boundaries for when we will and will not use AI. For example, people who know me know that throughout my career, I have steadfastly refused to use ghostwriters – I write everything that has my name on it. I have continued this principle into the AI era – I won’t use AI to suggest topics, do research for me, organize my thoughts, create an outline, develop any content, or anything else. Perhaps this is old-fashioned, but I believe that it strengthens the quality of the content I put out there, partially due to the reasons shared above. I think that security professionals should resist the temptation that is AI for certain things – we will all benefit from it in the long run.
Related: The Agent Smith Dilemma: Governing Agentic Identities
Related: The Key to AI Software Security Isn’t Speed, It’s Maturity
