There is a dangerous phenomenon occurring in boardrooms across São Paulo and Rio de Janeiro. Faced with real and visible violence, budgets are readily approved for 'the definitive solution': next-gen cameras, remote monitoring, AI, and perimeter sensors. The system is installed. The dashboard turns green. The feeling of safety is immediate. But often, what was purchased wasn't protection; it was merely 'Security Theater.'
Technology can detect and alert. Protection also depends on what happens next.
The fallacy of 'heroic' technology
Many executives treat security as an IT issue: 'if we install software X, the problem is solved.' This is a fatal conceptual error.
Cameras, sensors and alarms can deter intrusion, record events, detect activity, and trigger alerts or automated actions. Their value still depends on placement, configuration, and the response that follows. Technology does not remove the need to decide who acts, when, and on what information.
Furthermore, technology is democratic: it evolves for those who protect and for those who attack. Organized crime in these cities is not amateur; they often know the system's blind spots as well as the installers do.
Sources:[1]
AI is not a crystal ball
Artificial intelligence is often sold as a crystal ball. It is not. AI can identify patterns, flag activity and support decisions while events unfold. Its output depends on the task, the data and the operating environment.
An alert does not, by itself, establish a person’s intent. A specialist can add local context and assess what to investigate or escalate, while recognizing the limits of human judgment too. The practical question is how the system and the people responsible for it work together: what the system flags, who assesses it, and what action is possible.
Sources:[2]
Where the money evaporates
In investigations and in the field, I have seen budgets drained by three recurring mistakes. The 'Myopic High-Tech': 4K cameras positioned where they only capture the top of a head or sun glare, beautiful images, zero tactical utility. The 'Penny Wise, Pound Foolish': spending millions on servers but economizing on the frontline team, a cap and sunglasses can defeat some facial recognition systems if there isn't a trained person to read the suspect's body language. Volume without intelligence: monitoring rooms that look like NASA, but where operators are drowning in false alarms, so that when the real alert sounds, it is treated as noise.
The strategic choice
In high-risk environments, technology without diagnosis is just an expensive tool for documenting tragedies. The true competitive differentiator isn't who has the newest gadget. It is who possesses Scenario Intelligence. Real security happens when technology stops being the protagonist and becomes what it should be: an amplifier of human competence.
If you are reviewing your security investments, ask the hard question: are you buying real protection, or are you just paying a premium to watch the replay of the incident in high resolution?