Imagine that you are at home, taking care of your children on a quiet afternoon, when suddenly the roar of a tactical unit breaks down your door. They point guns at you, handcuff you in front of the children, and accuse you of a crime committed in a place you have never visited. **
Thus began the nightmare of Angela Lipps, a woman whose life was dismantled because facial recognition software decided that her face “matched” that of a criminal.
This is not a dystopian sci-fi script; It is a reality that forces us to ask ourselves: are we delegating our freedom to machines that do not know how to distinguish between data and a human being? **
Discover why AI, our greatest security promise, can become your worst executioner if we forget human supervision.
Angela Lipps’ ordeal: When the algorithm decides your destiny
Let’s do an imagination exercise for a moment: Overnight, your identity is replaced by a “criminal” label in a database. This happened to Angela Lipps in Tennessee. **
Without ever having set foot in North Dakota, facial recognition software analyzed a bank’s cameras and “decided” that she was the scammer they were looking for.
The most terrifying thing was not the machine error, but human apathy: an agent looked at his social networks, noticed a similarity in hairstyle and, without a single physical proof, signed his arrest. From there, Angela’s life fell apart.
She spent 108 days in a cell, treated as a dangerous fugitive, while the system ignored her pleas. In the end, proving his innocence was very simple: his Uber Eats receipts and his cell phone’s GPS confirmed that he was thousands of kilometers from the crime.
But by the time the justice system admitted the mistake on Christmas Eve, Angela had already lost everything: her house, her car and even her dog.
AI vs human limitations
It’s natural to wonder: If they fail so badly, why do police departments insist on using security technologies? The answer lies in our biological limitations.
As humans, we are imperfect: we get tired, lose focus, and are full of unconscious biases. Our memory is fragile and our ability to process thousands of faces in seconds is simply nonexistent.
That’s where AI appears as an irresistible promise. In theory, an algorithm doesn’t blink, doesn’t time out, and can track a digital “needle in a haystack” in record time. The problem arises when we confuse “efficiency” with “infallibility.” **
The case of Angela Lipps shows us that we have fallen into automation bias: that dangerous tendency to believe that if a machine says it, it must be true.
By trying to make up for our shortcomings with technology, we have created a system where surveillance is total, but human verification is conspicuous by its absence.
The two faces of surveillance: a dangerous balance
Using AI in security has implications that can save lives or, as we saw with Angela, destroy them. Let’s segment this impact to understand what is at stake:
The positive side: Efficiency and protection
- Speed of response: AI makes it possible to locate missing people or terrorist suspects in a matter of minutes, something impossible for a human team.
- Fatigue reduction: Prevents errors due to fatigue or omission from going unnoticed in 24-hour control centers.
The dark side: The erosion of rights
- Algorithmic biases: Many systems fail more with women and ethnic minorities, becoming discriminatory tools.
- Reversal of the burden of proof: You are no longer “innocent until proven guilty”, but guilty until you manage to disprove what the software predicted.
This imbalance creates a dehumanized justice where the code outweighs reality.
Intelligent security: technology at the service of the common good
To avoid tragedies like Angela’s, the next generation of technologies must include “explainable AI” models, capable of justifying why they have made a decision, allowing a human to audit the process before acting.
We will see mandatory verification systems that, instead of making final judgments, present lists of probabilities that require additional physical evidence to be validated.
Additionally, the development of trained algorithms with diverse data free of racial or gender bias will be essential to ensure equity.
Finally, the objective is for technology to act as a co-pilot that enhances our ability to protect, but for human judgment, empathy and legal responsibility to have the last word on a person’s freedom.
The human factor: our last line of defense
What happened to Angela Lipps was not just a code failure, it was a renunciation of our ability to question. The most advanced technology in the world is of no use if common sense is left out of the control room.
True security is not born from infallible algorithms, but from institutions that understand that behind every piece of data there is a life that can be destroyed in a click.
At the end of the day, you and I must demand that innovation never be an excuse for laziness. Because if we allow the machine to be judge and judge, we are all one upgrade away from becoming the system’s next bug.
This post is also available in: