Ancova interactions posthoc

Hi everybody,
I have a rather theoretical question about post hoc test in ancova with interactions.
Suppose I have situation where I analyse the impact of x on y in 3 different groups (factor: group), like in the picture below.
To do stats I use R.
So I run a lm( y~ x * group) than I do car::Anova()) to test if the interaction is significant, and it IS.
Than I do the post hoc for interaction using lstrends() from emmeans package. The result is that the slope for G3 is significanlty different from G1 and G2, and there is no difference between slopes of G1 and G2.

Now my question is, can I (if yes, than how) test for the difference of intercept between G1 and G2? Can I simply run ancova without interactions only for this two groups?

Thanks for any help.