The device itself could perhaps be better shielded, before implantation.

I think (I'm really no expert on this stuff) I've read that device itself is located just below the skull, but is connected to a relatively long and very thin electrodes (or electrode arrays) that go deep inside the brain matter.

I could be wrong, but if that's how it is - well, it's not trivial to shield a needle to prevent it acting as an antenna.

Would be great if someone who knows this stuff can tell us how it really is.

The brain is basically water and radio waves don't penetrate water.

You only need to shield the exposed parts.

It's possible that better shielding on the device would be more of a curse than a blessing. It would increase device size, and possibly cause absorption of the waves that would heat it - leading to serious brain damage.

Sure, but what would act as the stable ground for the shielding? You need to avoid anything physical crossing the skin/skull/etc. to avoid cleanliness issues, remember.

Do brain implants get firmware updates and such wirelessly?

My company produces brain implants. Technically you can upgrade the firmware wirelessly (we do this in testing all the time) but nobody wants to take the responsibility for doing that with patients. With more RAM available I think wireless upgrade will be the standard in a few years.

Wireless updates obviously have major security issues, but I wonder if there's a compromise that doesn't require surgery. Induction, perhaps? Updates transmitted through a wired induction pad pressed against the skin above the device?

Could you speak to the reason there isn't any mention of shielding, or why to layman that seems like the best approach but actually isn't?

I don't know the details but the biggest problem seems to be the interaction of the pulse generator with the leads. You can't shield that away. It's more complex.

Wireless updates viewed as riskier than surgery? Or is it just that the liability moves from your company to the surgeon?

For a wireless update you need to be able to store the new firmware in a temp area, verify it and then activate it. Old devices don't have enough memory but that's changing. Also, most new firmware needs new hardware features so this capability may not be that useful in practice.

It's more than liability, reducing physical access requirements opens up a much larger attack surface.

I know if someone's opened up my skull. I assume everyone has some sort of multi-band wireless transceiver in their pocket, because that's a cell phone.

No, they do not. Any device upgrades (including e.g. firmware) require surgery.

Well you could probably disguise it inside of one of those driver caps, and I would think a copper mesh might be better... Though depending on the source of the interference a full head (ski)mask-like hat might be required.

Well, that would be more like a Faraday cage (and of course it already exists as a Commercial product). Google for "Balaclava Shield", you will find exactly what you described.