In almost any operation to protect a country or its civilians, somebody is likely to die. A police officer. A civilian who happened to be there. The person on the other side of it. All three outcomes are currently treated as the cost of the work.
Karan starts from the position that they are not the same thing as the work.
Zero incidents is not achievable and pretending otherwise is dishonest. Zero deaths is a different target, and it is the one worth designing against. A target you will miss, held anyway, because the alternative is treating the count as a cost of doing business.
If you work inside a force or a ministry and this reads as naive, that is a conversation worth having directly.