DISHONEST customers across Spain are increasingly using AI to do a runner in restaurants. 

One of the cases was reported in 2025, when a couple of tourists in Tarragona showed them a picture of a payment receipt exceeding €4,000 . 

The employees examined the screenshot closely and wanted to confirm that they had indeed received the money before letting the couple leave.

Upon discovering they hadn’t been paid, they alerted the Mossos d’Esquadra (Catalan police), who arrested the two tourists after verifying that the receipt had been generated using AI . 

READ MORE: British tourist among slew of hotel guests robbed by crooked employee using room key cards in Spain’s Mallorca

Another notorious incident took place at a restaurant in Galicia, reported anonymously by a waiter on the ‘SoyCamarero’ (‘IAmAWaiter’) X account. 

At their restaurant, a British couple ordered several dishes, seafood and wine, and at the end they showed the employees a video of black insects running all over their desk. 

The customers threatened to share the video on social media if they did not let them leave, but refused to send them the videos when they asked to see them better. 

READ MORE: British tourist attacks two taxi drivers outside Mallorca’s Palma airport

Although the logo of the restaurant was printed onto the tablecloth, the staff eventually found some inconsistencies in the footage and asked them to pay.

The anonymous waiter believes that this type of scams are going to ‘become all the rage.’

The National Institute for Cybersecurity (INCIBE) has posted a list with a couple of hints to identify fake content generated by AI:

1. Verify the source : Make sure the content comes from a reliable or reputable source. Be wary of news from websites or profiles you don’t know. Check their reputation and see if other trusted media outlets or users are reporting the same information.
2. Analyze the quality of the content : pay attention to certain details to uncover them based on the type of content:
   1. Text:
      1. Coherence and grammar : Check if the text has grammatical errors or inconsistencies that are not common in articles written by humans. 
      2. Video:
         1. Blinking and facial movements : Deepfakes often exhibit an abnormal number of blinks or unnatural facial movements. 
         2. Incongruities between face and body : observe if there are mismatches between facial expression and body movements. 
         3. Sound quality : the sound may not match the image, showing offsets or variations in tone that do not correspond to the scene. 
      3. Audio:
         1. Tone and rhythm : Analyze whether the tone of voice and speech rhythm are consistent. Voice clones may exhibit unusual variations. 
         2. Background noise : The lack of background noise or the presence of strange sounds may indicate manipulation. 
         3. Speech context : Check if the content of the speech is consistent with what the person would normally say or in the context in which they usually speak.
      4. Image:
         1. Fine details : they may have flaws in certain details, such as hands with disproportionate fingers or backgrounds that are not aligned correctly. 
         2. Lighting and shadows : Check if the lighting and shadows are consistent throughout the image.

Click here to read more Food & Drink News from The Olive Press.