The Food and Drug Administration is expanding its regulation of the Internet.

It’s latest incursion? FDA is taking on a drug company’s use of the “Like” button on Facebook.

I wrote last week about a warning letter that FDA issued to a drug maker over that company’s use of a search bar on its web site. FDA faulted the firm for improperly linking disease queries directly with drugs that the company marketed.

Now FDA has taken action against a dietary supplement maker that used Facebook to “Like” an unapproved claim regarding one of its products.

Taken together, the warning letters show that FDA is ramping up its regulatory activity over the use of Internet tools by regulated companies.

The FDA has been pressed to issue explicit guidance on how it intends to regulate Internet advertising (and social media in particular). But the agency has resisted. It seems that FDA’s policy on Internet promotion is going to unfold in the same manner that the agency has traditionally laid out its regulatory policy over advertising and marketing — piecemeal, one warning letter at a time.

FDA has historically been reluctant to publish clear guidance when it comes to how it regulates promotion. The agency worries that writing down its policies on paper could invite legal challenges to its authority over the underlying commercial speech, which courts have often recognized is protected under the First Amendment.

But one must ask whether this purposeful ambiguity is any way to run a regulatory process. Without clear rules, we are likely to see companies inadvertently ensnared by FDA’s evolving policy over the Internet, until a clear mosaic emerges (from individual warning letters) around how the agency intends to exercise its oversight.

Like the issue surrounding the Internet search, FDA’s Facebook foray was first spotted by Alec Gaffney, news editor of Regulatory Focus, in an excellent article he posted to the RAPS website. You can follow Alec on Twitter at @AlecGaffney.

In its latest Warning Letter, FDA cited a multitude of serious violations against the marketer of a drug product called Poly-MVA. The company had sold its product through the use of testimonials that were not substantiated by good science, according to FDA. One seemingly blatant abuse, still found on the company’s website when Gaffney checked, was a testimonial from a “Mr. Doug Wray,” whose improper endorsement suggests that by taking Poly-MVA, he was cured of his multiple myeloma.

In the Warning Letter, FDA also made a novel mention of the company’s Facebook “Like” of one of these exaggerated testimonials. Until now, it was never clear whether FDA would treat a “Like” (or a re-tweet for that matter) as an endorsement of the underlying content. That question seems to have been answered.

FDA writes: “We also note claims made on your Facebook account… The following are examples of the claims: In a March 10, 2011 post which was “liked” by “Poly Mva.” That post read: “PolyMVA has done wonders for me. I take it intravenously 2x a week and it has helped me tremendously. It enabled me to keep cancer at bay without the use of chemo and radiation…Thank you AMARC”

The offending post has since been taken down. But FDA’s interpretation that a “Like” implies endorsement could be a precedent-setting action, says Gaffney.

The same policy could be extended to other uses of social media tools, making companies weary of engaging at all. For example, will companies be held liable for social content generated by independent users of their sites? Under the emerging outlines of FDA’s policy, it seems that FDA would hold sponsors accountable for all the content on their sites, including content generated by independent users.

But who knows for sure? And there’s the rub.

All of this would be better understood if FDA was willing to define a policy, publish it, and defend it. With clear rules, good actors would stay in line.

But FDA may be on shaky legal ground when it comes to how far it wants to extend its oversight onto these Internet tools. So the agency will prefer to trickle out its policy slowly, one Warning Letter at a time. Eventually, a clear pattern will emerge. Moreover, by embedding these novel Internet warnings into letters aimed at reckless firms that have broken other rules (and won’t evoke much sympathy) the FDA is likely to avoid challenges from the recipients of the letters, or from outside observers. The dietary supplement makers that FDA has targeted with these Internet warnings make easy targets.

In the meantime, the individual warnings start to shape a more comprehensive framework. FDA gets to make its policy and never even had to stake out what its principles are, or defend its right to regulate these areas in the first place.