In the wake of the epidemic of forgeries that have made their way into our database about ancient Israel, Ed Cook puts it well: "It is not an overstatement to say that biblical archaeology may require a generation of disciplined, rigorous re-examination of all unprovenanced epigraphic material in order to be regarded again as a scientific discipline."
The basic problem is 1) it is not hard to forge ancient inscriptions. Some knowledge of ancient languages and a half-decent forgery lab, of which there are apparently several, is enough. 2) There are massive (6 or 7 figures) rewards for doing it. 3) It is very hard to conclusively prove forgery, partly because of the way that media and scholarly debates work. They form around an adversarial, "on the one hand, on the other hand" approach. As the "debates" around the health effects of smoking and the certaintly of global warming show, it is always possible to get an expert to argue your point of view. Authority in the modern media environment can be very diffuse.
This is something that my fellow Hopkins grad Christopher Rollston, one of the most skilled and thoughtful epigraphers around, has been thinking about for some time. He's come up with the best methodology I've seen for dealing with it.
One take-home point: always, always say whether your text was excavated or not, and always ask. It can be pretty disturbing to read stuff by good scholars where they don't, or where they shrug their shoulders.
For more, see his “Non-Provenanced Epigraphs I: Pillaged Antiquities, Northwest Semitic Forgeries, and Protocols for Laboratory Tests” in Maarav 10, and “Non-Provenanced Epigraphs II: The Status of Non-Provenanced Epigraphs within the Broader Corpus” (forthcoming in Maarav 11).
Next: why this may not really be the problem at all.