John Kelly says Trump praised 'Hitler's generals' for loyalty

Contributing sentiments towards policy:

54% : "No, no, no, they were totally loyal to him," Trump said, according to Goldberg's telling of Baker and Glasser's reporting.
45% : In the interview with editor-in-chief Jeffrey Goldberg, Kelly affirmed reporting from Peter Baker and Susan Glasser's book, "The Divider: Trump in the White House," which quoted Trump asking Kelly, his chief of staff at the time, "Why can't you be like the German generals?"
45% : Kelly has confirmed publicly reports of similar exchanges that he had with Trump, when serving in his administration.
43% : Goldberg wrote that Kelly "told me that when Trump raised the subject of 'German generals,' Kelly responded by asking, "'Do you mean [Otto von] Bismarck's generals?'"Kelly continued, in the interview with Goldberg this week: "I mean, I knew he didn't know who Bismarck was, or about the Franco-Prussian War.
42% : "I need the kind of generals that Hitler had," Trump reportedly said in a private conversation in the White House, according to Goldberg, who cited "two people who heard him say this.""People who were totally loyal to him, that follow orders," Trump added, according to Goldberg.
34% : Goldberg reported that Trump has grown "more and more interested in the advantages of dictatorship," and he included other examples that he thought would demonstrate this alleged interest.
17% : "According to Goldberg, Kelly said Trump was not acquainted with Rommel.
15% : "Kelly also famously confirmed Trump called dead U.S. servicemembers "suckers" in 2018.
5% : According to Goldberg's account of Baker and Glasser's reporting, Kelly responded by explaining to Trump that the German generals "tried to kill Hitler three times and almost pulled it off," but Trump reportedly was not swayed by the correction.
3% : Trump also pushed back on that comment at the time, saying Kelly was "lying" and suggesting Kelly was motivated by "his hatred of me."

*Our bias meter rating uses data science including sentiment analysis, machine learning and our proprietary algorithm for determining biases in news articles. Bias scores are on a scale of -100% to 100% with higher negative scores being more liberal and higher positive scores being more conservative, and 0% being neutral. The rating is an independent analysis and is not affiliated nor sponsored by the news source or any other organization.