Bava Metzia 41B

ואני אומר אינה משונה.

R' Yoḥanan continues: And I say that misappropriation by a paid bailee is not different.

ומאי משונה? לא תאמר שליחות יד בשומר שכר, ותיתי משומר חנם: ומה שומר חנם שפטור בגנבה ואבדה שלח בה יד ⁠– חייב, שומר שכר שחייב בגנבה ואבידה לא כל שכן. למאי הלכתא כתבינהו רחמנא? לומר לך: שליחות יד אינה צריכה חסרון.

The Talmud elaborates: And what is meant by: Misappropriation that is stated with regard to a paid bailee is different from misappropriation that is stated with regard to an unpaid bailee? As one could claim: Let misappropriation not be stated with regard to a paid bailee, and derive it from misappropriation with regard to an unpaid bailee by means of an a fortiori inference: And if an unpaid bailee, who is exempt in cases where he claims theft and loss, misappropriated the deposit, he is liable to pay, then a paid bailee, who is liable in cases where he claims theft and loss, all the more so is it not clear that he is liable if he misappropriated the deposit? With regard to what halakha did God write misappropriation in the case of a paid bailee? It is to say to you: Misappropriation does not require loss; intent to misappropriate is enough to render him liable to pay.

ואני אומר: אינה משונה כרבי אלעזר, דאמר: דא ודא אחת היא. מאי דא ודא אחת? משום דאיכא למפרך: מה לשומר חנם, שכן משלם תשלומי כפל בטוען טענת גנב.

R' Yoḥanan stated: And I say that it is not different, in accordance with the opinion of R' Elazar, who says: This case and that case are one. The Talmud elaborates: What is the meaning of: This and that are one? It means that it was necessary to teach misappropriation in both cases due to the fact that it can be refuted by an a fortiori inference: What is notable about an unpaid bailee? He is notable in that he pays the double payment when he falsely states the claim that a thief stole the deposit. A paid bailee reimburses the owner only for the cost of the deposit in that case. The legal status of the paid bailee is not consistently more stringent than that of an unpaid bailee, and therefore no a fortiori inference is possible.