Because the Cauchy distribution has a finite variance but infinite mean.

Because the Law of Large Numbers requires a finite first moment (mean), which the Cauchy distribution lacks.

Because the sample size is always too small for the CLT to apply.

Because the distribution is not symmetric.