Having been looking into scientifically plausible forms of telepathy, I stumbled across the electroreception senses that exist in some electric fish (and in a few land-dwelling marsupials). My question is whether it's feasible for a human-like species to evolve the same kind of thing, and what limitations would be imposed by living on land where there wouldn't be a consistent source of water to be used as a conductor as for the electric fishes.
Your main problem is that water conducts electricity really well, whereas air conducts electricity really poorly - air is effectively an insulator. Otherwise you wouldn't need to touch a live electrical wire to be electrocuted, you would just have to stand nearby. (Although large voltages can cause sparks to jump a gap, but we're talking machinery not biology in those cases).
Therefore the distance over which your humans could sense anything with electrosenses is so short as to be useless. Years back I read some papers about rattlesnakes - someone suggested that the rattling tail could generate electrostatic charges which the snake could use to sense things with its tongue. But another scientist calculated the effective range and it was only a couple of centimetres. The snake would be able to see and smell its prey long before it could electrosense it. Sorry, this is the only reference I can find on the discussion.
So the Vulcan Mind Meld is possible - you touch the person. Knowing what someone across the street is thinking is not feasible.
Magnetoreception might be easier. You might be interested in magnetic implants which enable you to feel not only magnetic fields, but also (alternate) electric currents, and do not need a conducting environment. The nice thing is that your brain soon internalizes the new feeling and it looks like a new, separate sense.
You can even use non-invasive magnetic rings.
So, it is at least remotely plausible that evolution can lead to (magnetized) magnetite concentration in some part of the body - but that's still far from telepathy.
Check out the "aquatic ape hypothesis". It is plausible that humanoids evolve with water being an essential part of their lives and still live on land. In this case, something like electroreception might evolve given the right circumstances of course - there need to be electrical fields for starters. I hope I read your question correctly, it could be interpreted as "my humanoids live in the desert with almost no water" - in which case it is kind of obsolete but not really. Just ignore all the water bits. Yes, some land animals on earth have that ability, but there must be a reason why we didn't evolve that sense - so I conclude that it wasn't enough stimulation/need for it. But also note that you can sometimes feel electrical fields, when all your body hair starts to spread out because your body is charged - you can feel that actually.
Here is what I would do, but I'm no evolutionary biologist. I hope this would be plausible to enough readers:
There is this organism (say a plant) that is poisonous and produces a weak electrical field. There is another plant/fish whatever that copies that electrical field to scare off predators. Your humanoids and their ancestors love to eat that organism. Variations are countless, from the initial plant not being poisonous to the humanoids to food growing where there are natural electrical fields (different question how they might get there) and so on.
I think one should have this in a more aquatic setting. Yes, there is no reason why this wouldn't work in air, but I think your question has a bit of a "what is the most likely way" quality to it.
The problem with non-aquatic electrical fields of course is that they need to be stronger, maybe resulting in regular shock. If you go for the desert option with little water, maybe choose an environment where not getting electrocuted is essential. How to construct this planet is a different question.
You could imagine a completely different reason for electric fields on your planet. For example your humanoids needed to go someplace for seasonal change and that is made obvious by regular strong solar winds that charge the atmosphere or strong storms. The options really are limitless, but I still feel the most intuitive way is having your land-living humanoids be in water a lot which is even how we might have evolved.
What the fish you mention do is likely more akin to an iPhone screen, they send preferably alternating current through a conductive medium and recognize any changes in resistance and capacitance. This is mostly not viable unless you share a circuit with the thing you hope to detect. Nevertheless, some touch devices can sense your fingers through air, though I don't know what that kind of accuracy requires. That failing, you can use induction to detect charge without a circuit but not capacitance. You wouldn't be easily able to discern between distance and magnitude without two sensors.
If you also hope to detect the presence of charge, here are some options. Be aware that the charge of neurons would interfere and is highly dynamic.
Introduction to the vicinity of charge will induce a motion of electrons through certain materials. This is why our hair becomes positively charged near lightning, and is an event that facilitates the jump of electrons from the clouds. It works at a small scale incredibly well. Look up "induction".
Changes in charge will also induce magnetic fields that can then affect the motion of charge within them.
I suspect that charge is far easier for relevant biological life to maintain than a solid state magnetic field, because most magnetization is easily lost. Lodestones are made by lightning and wouldn't be the most common thing in the world, yet every animal creates charge to move neurotransmitters through and between neurons. Baking destroys most high quality magnets like those used in motors. Electromagnetism might even be more viable than synthesizing one for some applications. You could have magnetite seeking hermit crabs though.