The nucleosynthesis production of fluorine (F) is still a matter of debate. Asymptotic giant branch (AGB) stars are one of the main candidates for F production. However, their contribution to the total F budget is not fully known due to the lack of observations. In this paper, we report the detection of aluminium monofluoride (AlF) line emission, one of the two main carriers of F in the gas-phase in the outflow of evolved stars, towards five nearby oxygen-rich (M-type) AGB stars. We studied the Atacama large millimetre/sub-millimetre array (ALMA) observations of AlF (v = 0, J = 4—3, 9–8, 10–9, and 15–14) and (v = 1, J = 7–6) line emission towards o Ceti, and (v = 0, J = 7–6 and 15–14) lines towards R Leo. We also report a tentative detection of AlF (v = 0, J = 7–6) line in IK Tau, (v = 0, J = 15–14) line towards R Dor, and (v = 0, J = 7–6 and J = 15–14) lines in W Hya. From spatially resolved observations, we estimated the AlF emitting region with a radius ~11R_⋆ for o Ceti and ~9R_⋆ for R Leo. From population diagram analysis, we report the AlF column densities of ~5.8 × 10¹⁵ cm⁻² and ~3 × 10¹⁵ cm⁻² for o Ceti and R Leo, respectively, within these regions. For o Ceti, we used the C¹⁸O (v = 0, J = 3–2) observations to estimate the H₂ column density of the emitting region. We found a fractional abundance of f_AIF/H2 ~ (2.5 ± 1.7) × 10⁻⁸. This gives a lower limit on the F budget in o Ceti and is compatible with the solar F budget f_F/H2 = (5 ± 2) × 10⁻⁸. For R Leo, a fractional abundance f_AIF/H2 = (1.2 ± 0.5) × 10⁻⁸ is estimated. For other sources, we cannot precisely determine the emitting region based on the available data. Assuming an emitting region with a radius of ~11R_⋆ and the rotational temperatures derived for o Ceti and R Leo, we crudely approximated the AlF column density to be ~(1.2−1.5) × 10¹⁵ cm⁻² in W Hya, ~(2.5−3.0) × 10¹⁴ cm⁻² in R Dor, and ~(0.6−1.0) × 10¹⁶ cm⁻² in IK Tau. These result in fractional abundances within a range of f_AIF/H2 ~ (0.1 − 4) × 10⁻⁸ in W Hya, R Dor, and IK Tau.