Unarmed strikes are negated by the immunities listed

According to the errata for the monster manual, the immunities in question have been changed from immunity to weapons to immunity to attacks. The exact quote is:

Throughout the book, instances of “nonmagical weapons” in Damage Resistances/Immunities entries have been replaced with “nonmagical attacks.”

Note that this had always been the case. Before any errata, unarmed strikes were listed as weapons, and the immunities referenced the weapons. Unarmed strikes were corrected to not be weapons, and the immunities were corrected to reference attacks instead of weapons.

This is a kind of musing, rambling answer because there are a few aspects of the rules that I think deserve calling out and considering here. There is no explicit rule on this one way or the other, and I haven’t found any commentary by Paizo, so I am going to have to reason about what the rules do explicitly say to inform an answer. (I am not aware of any examples of this, or commentary on this, from any edition of D&D, but that doesn’t necessarily mean there isn’t one; that would be a lot of material to sift through to rule out the possibility.)

1. The lich’s phylactery is probably magical, but it’s not actually called out as such.

   Basically, nothing ever calls it a magic item. You use Craft Wondrous Item to make it, and that feat generally creates magic items (but the phylactery could be an exception), and its crafting rules require a minimum caster level (but strictly speaking that only says you have to be magical). So there is good reason to assume it is magical, but it’s not spelled out.

2. It doesn’t necessarily matter if the phylactery is magical.

   The lich’s rejuvenation ability is a property of the lich, not the phylactery, though obviously they interact. But ultimately, the only thing that the rejuvenation ability requires of the phylactery is that it be the phylactery—and it’s not clear that a phylactery, even if it is normally magical, ceases to be a phylactery when magic is suppressed.

3. Dispel magic probably doesn’t come into play.

   The lich’s rejuvenation is a supernatural abilities, and those cannot be dispelled. You can argue that you could cast dispel magic on the phylactery, if you have decided it is magical in the first place, and then that the phylactery ceases to be a phylactery while suppressed, but otherwise dispel magic is of no use.

4. A lich who is destroyed in an antimagic field has rejuvenation suppressed.

   This one is pretty clear: suppressing magic around the lich suppresses its magical abilities. Since supernatural abilities are magical, that includes rejuvenation.

   The main question is whether or not the lich only gets one chance at rejuvenation. Supernatural abilities suppressed by antimagic field come back as soon as the antimagic field goes away—but at that point, the lich is already destroyed. Can a destroyed undead creature even be said to have any abilities?

   Moreover, even if it does, the rejuvenation ability is “triggered” by a particular event, the destruction of the lich. It specifically says “When a lich is destroyed,” not something like “A destroyed lich...” So by suppressing the ability at the moment of that event, does the triggered ability just miss the trigger? Or does it trigger, but get suppressed until it can go?

5. If a lich is destroyed while its phylactery (but not it itself) is in an antimagic field, then... unclear.

   This is basically the same as the question of dispelling the phylactery: can rejuvenation still target it, that is, is it still a phylactery without magic? First aspect of that question is whether or not the phylactery was ever magical to begin with, as discussed, and even if so, whether that was a required aspect of its identity as the lich’s phylactery. Either answer gives rise to more questions.

   If the lich’s phylactery wasn’t at the time of its destruction, does that mean it missed its window to rejuvenate? Or does the ability trigger, only to find its target suppressed and so become effectively suppressed itself, allowing it to resume once the antimagic field is over? Or does the triggered ability just fall apart because the phylactery wasn’t available at the time?

   On the other hand, if the lich’s phylactery is still a phylactery even though it is suppressed, doesn’t that mean that the lich—in spirit, anyway—moves adjacent to it and thus is now inside the antimagic field? If that were the case, we would have a pretty clear case of the ability being suppressed—after being successfully triggered. That would mean that the lich doesn’t begin rejuvenating until the antimagic field is gone, but then does begin to rejuvenate. And if the lich isn’t said to move yet, and suppressing the magic of the phylactery itself doesn’t prevent rejuvenation, then you would have the lich magically reconstructed while within an antimagic field—decidedly odd, but conceivable, at least under the rules.

These questions basically give rise to a series of options:

• If the phylactery is not magical

• Or the phylactery’s magic is not integral to its identity as phylactery

  Then

  • Dispel magic doesn’t matter

  • Antimagic field on the phylactery may matter, specifically:

    • If the lich moves to the phylactery as soon as rejuvenation begins

      Then the rejuvenation is suppressed until the antimagic field ends but resumes at that point.

    • Otherwise the rejuvenation happens unimpeded and the lich’s magic is only suppressed once it is complete.

• Otherwise (if the phylactery is magical and must stay magical)

  Then

  • Dispel magic and antimagic field on the phylactery have the same effect as antimagic field on the lich.

• Either way antimagic field on the lich suppresses rejuvenation, meaning:

  • If the trigger is a one-shot deal

    Then the lich does not rejuvenate and is dead.

  • Otherwise the lich rejuvenates once the antimagic field around him (or around the phylactery, or dispel magic on the phylactery, if those are magical and must be so to rejuvenate) goes away.

Since the phylactery is probably magical, and seems to be doing magical things (“phylactery [...] begins to rebuild the undead spellcaster’s body nearby” implies action by the phylactery which implies it is magical), I tend to favor that suppressing magic on it is much like suppressing the lich’s own magic. After all, for the most part, if you have unimpeded access to a lich’s phylactery then you can probably destroy it anyway, so it doesn’t make a whole lot of difference and seems more appropriate that way.

The real question, to me, is whether suppressing rejuvenation at the time of the lich’s destruction prevents the trigger altogether, or if it only delays it until the magic-suppression ends or is removed. Rules-as-written, it would seem to me that it’s a one-shot deal, based on the word “when.” If they had written it with “if” instead (as in, “if the lich is destroyed”), then I would go the opposite way on that. But the rules-as-written approach here seems somewhat dubious to me, on the basis that we are hanging a lot of weight on the choice of “when” over “if” for a ruleset that is not generally written with that level of care.

And narratively, I like the idea that the suppression is only good for as long as you can maintain it. It seems interesting to me to imagine having to set up some system to keep magic suppressed in the area—and an obvious future plot-hook, to have that suppression removed by scheming fools or by nosy teenagers.

But I could also see wanting to give this as an option for finishing a campaign, so that you have this epic battle wherein you have to try to make sure the lich is finished while inside the antimagic field (which would be fiendishly difficult) or else it will just reform (assuming here that for plot reasons the phylactery is inaccessible or undiscoverable).

So I think in large part it depends on how you want to run your campaign.