Impacts due to near-Earth objects (NEOs) are responsible for causing some of the great mass extinctions on Earth. While nearly all NEOs of diameter > 1 km, capable of causing a global climatic disaster, have been discovered and have negligible chance of impacting in the near future, we are far from completion in our effort to detect and characterize smaller objects. In an effort to test our preparedness to respond to a potential NEO impact threat, we conducted a community-led global planetary defense exercise with support from the NASA Planetary Defense Coordination Office. The target of our exercise was 2012 TC4, the 10 m diameter asteroid that made a close pass by the Earth on 2017 October 12 at a distance of about 50,000 km. The goal of the TC4 observing campaign was to recover, track, and characterize 2012 TC4 as a hypothetical impactor in order to exercise the global planetary defense system involving observations, modeling, prediction, and communication. We made three attempts with the Very Large Telescope (VLT) on 2017 July 27, 31 and on 2017 August 5 and recovered 2012 TC4 within its ephemeris uncertainty at 2.2 arcmin from the nominal prediction. At visual magnitude V = 27, the recovery of 2012 TC4 is the faintest NEA detection thus far. If an impact during the 2017 close approach had been possible based on the 2012 astrometric data, these recovery observations would have been sufficient to confirm or rule out the impact. The first automatic detection by a survey (Pan-STARRS1) was on September 25, which is the earliest that 2012 TC4 would have been discovered in survey mode, if it had not been discovered in 2012. We characterized 2012 TC4 using photometry, spectroscopy and radar techniques. Based on photometric observations, we determined a rotation period of 12.2 min with an amplitude of 0.9 magnitudes. An additional lower amplitude period was detected, indicating that 2012 TC4 was in a state of non-principal axis rotation. The combined visible and near-infrared spectrum puts it in the taxonomic X-class. Radar images at 1.875 m resolution placed only a few range pixels on the asteroid, reveal an angular, asymmetric, and elongated shape, and establish that 2012 TC4 is less than 20 m on its long axis. We estimate a circular polarization ratio of 0.57 + -0.08 that is relatively high among NEAs observed to date by radar. We also performed a probabilistic impact risk assessment exercise for hypothetical impactors based on the 2012 TC4 observing campaign. This exercise was performed as part of ongoing efforts to advance effective impact risk models and assessment processes for planetary defense. The 2012 TC4 close approach provided a valuable opportunity to test the application of these methods using realistically evolving observational data to define the modeling inputs. To this end, risk assessments were calculated at several epochs before and during the close approach, incorporating new information about 2012 TC4 as it became available. Two size ranges were assessed-one smaller size range (H = 26.7) similar to the actual 2012 TC4, and one larger size range (H = 21.9) to produce a greater-damage scenario for risk assessment. Across the epochs, we found that only irons caused significant damage for smaller size. For the larger size case, however, hydrous stones caused the greatest damage, anhydrous stones caused the least damage, and irons caused moderate damage. We note that the extent of damage depends on composition in different size regimes and, after astrometry, size is the most important physical property to determine for an incoming object.