New York CNN Business  — 

Twitter last year proposed a policy that would have brought in tougher rules on dehumanizing language on the platform. Amid their deliberations, Twitter’s top executives reportedly considered as an example Trump’s comments — made during a meeting — about “shithole countries.”

But Twitter decided not to adopt the sweeping proposal — and President Trump is still at it. On Saturday he attacked Democrat Elijah Cummings, whose congressional district includes Baltimore, and criticized the city. He said “no human being” would want to live there and said it was “disgusting, rat and rodent infested.”

Indeed, Twitter confirmed on Sunday that Trump’s tweets did not violate its rules at all.

Last September, Vijaya Gadde and Del Harvey, two senior Twitter employees, outlined a proposed “new policy to address dehumanizing language on Twitter.”

The proposal defined dehumanizing language as “Language that treats others as less than human,” adding, “Dehumanization can occur when others are denied of human qualities.”

Dehumanizing language aimed at a range of groups including religious and ethnic groups would break Twitter’s rules, the policy proposed. So too would language aimed at people in a specific location.

Mashable first pointed to the proposal in the context of Trump’s tweets about Baltimore.

In developing its new rules, Twitter’s top executives considered as an example Trump’s comments about “shithole countries,” The New York Times reported.

But when Twitter finally rolled out the new rules earlier this month, it was a scaled back version of what the company had proposed last September.

Twitter banned dehumanizing language against religious groups, as well as other protected categories — which include people of color and “marginalized and historically underrepresented communities.”

Dehumanizing language against people based on where they live would not be banned as originally proposed.

Twitter said they arrived at the decision after internal and external consultation.

“Respondents said that ‘identifiable groups’ was too broad, and they should be allowed to engage with political groups, hate groups, and other non-marginalized groups with this type of language,” Twitter said in a blog post.

“In other instances, people wanted to be able to refer to fans, friends and followers in endearing terms, such as ‘kittens’ and ‘monsters,’” it added.

Many observers argue that Trump’s criticism of Baltimore followed a trend of the president using language like “infestation” to criticize lawmakers who are people of color.

Twitter announced last month that instead of removing most tweets from world leaders that broke its rules, it would label and down-rank them.

But when the president tweeted racist language in his attack on four progressive Democratic congresswomen two weeks ago, Twitter did not label the tweets — an apparent contradiction of its own written policies.