There is a statistically significant difference.

The two means are definitely equal.

We fail to reject the null hypothesis of no difference at that confidence level.

The sample size was too small.