At the APTLD-74 meeting the issue about two-character new TLD came up.

Looking at the summary of the recommendations at https://gnso.icann.org/sites/default/files/file/field-file-attach/subsequent-procedures-initial-annex-c-02jul18-en.pdf we can find the proposal to remove the limitation about two character strings. The text reads:

2.7.1.c.3: The Work Track is also considering a proposal to remove the reservation of two-character strings at the top level that consist of one ASCII letter and one number (e.g., .O2 or .3M), but acknowledges that technical considerations may need to be taken into account on whether to lift the reservation requirements for those strings. In addition, some have expressed concern over two characters consisting of a number and an ASCII letter where the number closely resembles a letter (e.g., a “zero” looking like the letter “O” or the letter “L” in lowercase looking like the number “one”).

Personally, I believe that besides the issue about confusing similarity there is a more fundamental problem. Two characters string are used for ccTLDs following inclusion in the ISO 3166-1 Alpha-2 list.

While the list is clearly identified as “Alpha-2” I am not sure that, in case of a future shortage of combinations, ISO could not instruct ISO 3166-MA to use also two-characters alphanumeric strings. We have seen that already with IATA and the airline codes, initially limited to two characters alphabetic strings, but later extended to two characters alphanumeric.

I am always suspicious when something that is not necessary and potentially dangerous is proposed.