I've been giving some additional thought to the ethics issues associated with metadata, both because of the recently issued opinions in Arizona and Pennsylvania and because I've recently given several lectures on the problem.
In my view, I think it's impossible to develop a blanket rule that captures the complexities of this issue. So I'm inclined to like the Pennsylvania approach, which effectively says that a lawyer who receives an electronic document containing potentially privileged metadata should be able to look at that metadata, but only in certain circumstances. Although this "sometimes it's ok" approach needs to be fleshed out with concrete illustrations in order to ensure that lawyers have adequate guidance, I think that a nuanced approach is essential.
To see why, consider two examples drawn from litigation. In the first, a lawyer receives discovery documents from an adversary, some of which are in electronic form. Some of the electronic documents have been converted to .pdf, and some of the text in those documents has been digitally redacted through the use of what is effectively a digital black magic marker. In many cases, this digital redaction may cover the visible text, but a savvy user of electronic documents can easily discover the text that lies beneath the digital "black out." Should the recipient of such a document try to see what's underneath?
In my view, the answer should be no. The sender clearly thought that the underlying text was privileged, going so far as to make it unreadable upon opening the document. By extracting the text underneath the blacked out portions, the recipient would be using technology with the intent to learn information that the recipient knew to be privileged. Looking at this metadata seems akin to using some fancy listening device to overhear a conversation between an opponent and her lawyer, when that conversation was intended to be private. In both cases, the receiving lawyer is trying to uncover privileged information that the sending lawyer took reasonable steps to prevent from disclosure.
In contrast, consider a lawyer who receives an electronic document in its native format, say Microsoft Word, which might contain informative metadata, such as who originally authored the document or when it was created. Perhaps the "track changes" feature was turned on, so the receiving lawyer might also be able to see who worked on the document and what changes that person made to the document. Should the lawyer try to extract that information?
In my view, the answer to this question should be yes. If the document was discoverable, the information contained in the document's metadata will, in many cases, be discoverable as well. Given the current state of the law, the lawyer would be wise to get court approval before digging into the metadata (there are a few cases to this effect), but in general, I wouldn't be troubled to hear that a lawyer extracted this sort of information even without court approval. Although the review of the information might occasionally reveal privileged information, the circumstances of that revelation don't strike me as ethically problematic.
And then, of course, you get into documents that are transmitted outside of the litigation context, such as contracts or other kinds of transaction documents. I think we might reach different conclusions about the propriety of reviewing the metadata in these documents, depending on the circumstances.
That's a long way of saying that I think the trend in this area should be in the direction of a nuanced approach. I don't think a single answer -- either "never look at metadata" or "it's usually ok to look at metadata" -- makes a lot of sense. But we do need to give lawyers some guidance, so my hope is that future opinions not only recognize the nuance (like the Pennsylvania opinion did), but that they also describe some concrete examples and explain what lawyers should do in specific, commonly experienced situations.