As mentioned in the comments, why not? A boron-boron bond will form as long as there is an increase in stability of the system. Forming a bond requires energy, but there is a lot of energy released once it gets formed. The feasibility of a bond depends which energy is greater.
One example of where a boron-boron bond occurs is in Aluminium diboride / Magnesium diboride
Here, you have graphite-like "sheets" of boron (pink), with aluminium layers between them.
This will probably be quite stable, due to delocalization. The singly-filled aluminium $p$-orbitals will overlap with the empty $p$ orbitals of Boron, and will delocalize across the graphite-like sheet (I cannot confirm this, though). This will lead to a large increase in stability.
A few more compounds with $\ce{B-B}$ bonds are Lanthanum hexaboride and various Yttrium borides. These are probably very stable due to the crystal structure (though delocalization may be a factor as well)