Here's one simple way to look at it classically: Photons have energy, and by $m = E/c^2$, we know that energy means mass. So, photons are attracted by gravity just like ordinary matter. A black hole is simply a gravitational field that is so strong that photons can't reach escape velocity.
Now, from a quantum viewpoint, that means this: Any photon that makes it up to the event horizon will have zero energy left, and so has nothing left to tunnel through the horizon with. Or equivalently: While you are correct that photons can in principle tunnel through an event horizon from the inside out, if the only way they can reach that event horizon is by losing all their energy, the option of tunneling means nothing. So, an object within the black hole cannot send a photon out via tunneling.
But there is still one more way to interpret your question: Can pairs of virtual photons form at the event horizon, with one falling in (with negative energy) and the other escaping outwards (with positive energy)? The answer to that one is: Sure! In fact, I'm pretty sure that is the dominant mechanism by which Hawking radiation is produced in a quite smallish (e.g. dust-particle down to molecule sized) but not-quite-ready-to-go-boom black hole. The reason of course is that photons can be produced in pairs at any level of energy, whereas cases such as electron-positron pairs require a lot higher minimum level of energy to participate significantly in Hawking radiation events.
So, just as a hot piece of iron starts radiation first with photons, and has to get a lot hotter (understatement!) to start giving off matter-antimatter pair creation radiation, a black hole that is getting small enough to start radiating non-trivially will begin by giving of simple thermal black-body radiation (wow, that's almost like a pun, isn't it?).
Finally, notice the important distinction in my two answers: Photons traveling from within the black hole cannot tunnel through, and so cannot convey information about the interior. However, photons created with quantum randomness at the event surface, and thus not containing information about the interior (though there's been a lot more theorizing about this since Hawking first proposed his ideas) can radiate outward, but without providing specific information about the interior of the black hole.
Or so it is all theorized. Direct experimentation is, ah, difficult for this domain.