Theories of planet formation give contradicting results of how frequent close-in giant planets of intermediate mass stars (IMSs; $1.3\le M_{\star }\le 3.2\, \mathrm{M}_{\odot }$) are. Some theories predict a high rate of IMSs with close-in gas giants, while others predict a very low rate. Thus, determining the frequency of close-in giant planets of IMSs is an important test for theories of planet formation. We use the CoRoT survey to determine the absolute frequency of IMSs that harbour at least one close-in giant planet and compare it to that of solar-like stars. The CoRoT transit survey is ideal for this purpose, because of its completeness for gas-giant planets with orbital periods of less than 10 d and its large sample of main-sequence IMSs. We present a high precision radial velocity follow-up programme and conclude on 17 promising transit candidates of IMSs, observed with CoRoT. We report the detection of CoRoT-34b, a brown dwarf close to the hydrogen burning limit, orbiting a 1.1 Gyr A-type main-sequence star. We also confirm two inflated giant planets, CoRoT-35b, part of a possible planetary system around a metal-poor star, and CoRoT-36b on a misaligned orbit. We find that $0.12 \pm 0.10\, {{\ \rm per\ cent}}$ of IMSs between $1.3\le M_{\star }\le 1.6\, \mathrm{M}_{\odot }$ observed by CoRoT do harbour at least one close-in giant planet. This is significantly lower than the frequency ($0.70 \pm 0.16\, {{\ \rm per\ cent}}$) for solar-mass stars, as well as the frequency of IMSs harbouring long-period planets ($\sim 8\, {{\ \rm per\ cent}}$).