Yassine Maizi uses AI for research, not the final fact-check
Yassine Maizi is Founder & Editor of SafeStackPro, which reviews and compares security and AI tools for freelancers and small businesses without an IT department. The conversation looks at where AI saves time, why every published claim is checked against its source, and how the site is preparing for AI-driven discovery without assuming it is already bringing traffic.
AI can speed up repetitive research and site tasks, but responsibility for checking every quote, price and claim stays with a person. For SafeStackPro, publishing less and keeping reviews accurate is more valuable than chasing volume.
What’s the quick origin story of your business, and what makes what you do genuinely different from the other options your customers are looking at?
I started it because, as a freelancer myself, I struggled to find honest answers on which security tools were worth paying for. What I try to do differently is be honest about what I’ve actually used. If I only have the free plan of a tool, I say so in the review instead of pretending I ran lab tests on it. A lot of sites in this niche publish benchmark numbers nobody really measured. I’d rather publish less and have every claim hold up.
Where does most of your new business come from today, and where do you wish it came from?
Almost all of it comes from Google search, and honestly it’s still small. The site is young and I’m building authority slowly. Revenue comes from affiliate programs, so it depends entirely on readers finding a review and trusting it. I wish more of it came from recommendations: people sharing a comparison with a colleague, or newsletters and journalists citing our pages. That kind of traffic converts better because the trust is already there.
What’s one thing you’ve actually handed over to AI in the last year, and one thing you tried it on and went back to doing the human way?
Handed over: the technical and research grind. I use Claude for site audits, finding broken links, checking structured data, and digging through pricing pages, and Notion AI to keep my notes and drafts organised. That used to eat my evenings. Went back to human: fact-checking. Early on, AI-assisted drafts slipped in quotes and statistics that looked credible but didn’t exist. I ended up going back through dozens of published articles to remove them. Now every quote, price and claim gets checked against the original source by me before anything goes live. AI can draft, but it can’t be the one vouching for the facts.
Have you noticed a change in how customers find you or what they already know before they reach out? Is anyone arriving through AI assistants like ChatGPT yet?
Not in a way I can clearly measure yet. Search traffic is still the main channel. What I have changed is how I write: clear answers near the top, FAQs, accurate pricing, real sources. Those are the pages AI assistants tend to cite, so I’m treating it as preparation rather than something that’s already paying off.
If you were advising someone in your industry on all of this for the next 12 months, what would you tell them to actually do, and what would you tell them to ignore?
Do: use AI to save time on the repetitive work, but keep a human responsible for every fact you publish. One fabricated quote can cost you more trust than a hundred fast articles earn. Be upfront about what you’ve actually tested. Ignore: the pressure to publish at volume. I stick to about one article a week and spend the rest of the time fixing and updating what’s already live. Readers and search engines both reward pages that stay accurate.
Thank you to Yassine Maizi and the team at SafeStackPro for sharing what actually worked with Leaders Perception readers.
Want to share your playbook?
Leaders Perception publishes founder and operator interviews every week. A few questions, about five minutes, no calls. It is free and everyone we accept gets published.