In a paper posted recently to arxiv.org, physicist Roberto Casadio of the University of Bologna in Italy and his co-authors argue against such a scenario. But the bulk of the attention following their analysis has focused on their observation that microscopic LHC black holes, should they arise, could persist for seconds before decaying. (To wit, Fox News's story headlined: "Scientists Not So Sure 'Doomsday Machine' Won't Destroy World.")
It's worth pointing out that these collider-induced black holes only arise in certain theoretical frameworks, which posit that we reside in a universe of more dimensions than the four (three for space, one for time) that we're used to. In fact, the Casadio team's analysis presumes a certain five-dimensional theoretical model of the universe known as the Randall-Sundrum (RS) model. Although the RS model has its adherents, many alternatives exist—in fact, it is hoped that the LHC will help illuminate which of the competing big-picture theories is correct (or at least more likely to be correct).
It's also worth pointing out that such concerns have been raised before. As noted in a 2007 New Yorker feature about the LHC, a similar argument was raised in 1999 around the start-up of Brookhaven National Laboratory's Relativistic Heavy Ion Collider (RHIC). The article points out that Scientific American played a role in the controversy by printing a letter asking whether a black hole at Brookhaven could devour the Earth in minutes, along with a reply from physicist Frank Wilczek, then of the Institute for Advanced Study in Princeton, N.J. In his response, Wilczek dismissed the black-hole scenario but mused that an alternate disaster might theoretically arise via the production of strangelets, a hypothesized form of matter. Although he cautioned that such a catastrophe was "not plausible," the media seized on the mention and Brookhaven was forced to convene a panel of physicists to vouch for the collider's safety. (The RHIC has been running since 2000.)
Even presuming we live in a universe in which the LHC is capable of producing black holes, and even presuming that the initial conditions are such that the miniature gobblers can survive for some seconds, Casadio and colleagues calculate that such black holes would be unable to grow to catastrophic size before decaying. But, like Wilczek's strangelet comment, the extended black-hole lifetimes hypothesized by Casadio and his co-authors are bringing collider safety concerns back to the fore.
PHOTO OF REPLACEMENT FOR DAMAGED LHC MAGNET © CERN