We show that Borodin's conjecture on diagonal coloring is false by improving the bound conjectured to be tight.
