Interesting that the criticism is levelled at Autodesk 'the company' rather than that it admits what people don't want to say: Autocad is pretty good, and it will take a large and well funded entity to take it on. Using anti-trust tools may be the right tool here but the industry as a whole was quick to adapt to Autocad long before it became a dominant player.
Autocad - not Autodesk - has many 100's of years of development done to it. It was essentially game over by the mid 1990's and since then they have only further consolidated their position. BIM hookups and the very large amount of software that somehow interacts with Autocad or has been written in AutoLisp are the kind of lock-in that newly minted start-ups can only dream of.
Improving on Autocad is going to be very hard, it will take a long time and even then you will probably lose the game. Acquisition is probably your best bet, unseating a dominant player that doesn't make any major mistakes is super hard.
I'm not sure what can be done about it, other than forcing the opening up of all file formats and interoperability information. And lots of that is already open.