With version 10, I did this:

gammadist:=(x^(alpha-1)*exp(-x/theta))/(GAMMA(alpha)*theta^alpha);

simplify(int(gammadist,x=0..infinity),symbolic);

The result of the integral is 1.

When I do this exact same thing in version 12, I get an integral without any symbolic computation.

If I use assum(alpha>0): additionally(theta>0): then it will give me the same result.  However, in version 10 I also did

mgf:=simplify(int(exp(t*x)*gammadist,x=0..infinity),symbolic);