Why you can no longer trust your senses

Scams have existed since the dawn of humanity. Every technological innovation has been misused by fraudsters to make a profit. We only think about text messages or emails that want our passwords. With AI tools improving every day, the point has come when common sense is no longer enough. A commentary analysis.

AI fraud continues to increase

• The Federal Office for Information Security (BSI) issues an annual report on the situation of IT security in Germany. The latest analysis shows that more than one in five Germans (22 percent) have already been affected by cybercrime. In 2025, seven percent of Germans will be Victim of an online crime become.
• At 43 percent, “fraud” comes in first place, ahead of fake online shops (22 percent) and data theft (20 percent). AI applications end up relatively far down in the ranking at two percent. However, the existence of AI fraud and AI manipulation in an official evaluation underlines that artificial intelligence is also on the dark side of the internet has arrived.
• The risk barometer from the insurer Allianz, which was published in January 2026, also underlines how great the risk actually is. According to this, “artificial intelligence” is the 32 percent after cyber incidents second largest source of danger. Not just anyone shares this assessment, but over 3,000 risk management experts in a representative survey.

Biggest weak point: people

Allianz’s risk assessment with regard to the dangers posed by artificial intelligence is certainly not exaggerated. This is evident from the alone diverse possible uses for AI Crime. Perfect messages without spelling or grammatical errors and tailored to the victim’s personal writing style? Delivers AI.

Fake calls of supposed accidents in which supposed relatives cry into the phone? Can be easily created from audio files – for example from publicly viewable Instagram Reels. AI regulates.
Your boss whose AI image asks you to make a transfer from your company account via a deepfake in a Teams meeting? AI makes it possible.

The problem – professionally and privately – is ourselves People are and remain the greatest weak point in technical systems. We respond wonderfully to emotional blackmail and forget to think when pressed for time. As a result, we make mistakes that we would deny that we made in everyday life. Transfer money to a stranger? Never! Nevertheless, fraudsters steal large sums of money every day.

Voices

• BSI President Claudia Plattner and Federal Minister of the Interior Alexander Dobrindt In their conclusion, the population has a duty: “Protecting attack surfaces is the decisive lever for improving cybersecurity in 2026. As this year’s report shows, attackers are increasingly choosing easy-to-attack targets with poor resilience. Only those who actively protect themselves increase the chances of avoiding threats or minimizing harmful effects. Citizens must increase their awareness of cybersecurity and become more resilient.”
• The Internet Organized Crime Threat Assessment (IOCTA) of the European police authority Europol warns in a report: “Through the widespread use of Large Language Models (LLMs) and generative AI, social engineering methods are becoming more powerful: communication is tailored to the victims while criminal processes are automated at the same time.”
• Michael Bruch, Global Head of Risk Consulting Advisory Services at Allianz Commercialdescribes artificial intelligence as a big mover: “The rapid evolution and adoption of AI is fundamentally reshaping the global risk landscape, making it one of the key threats facing businesses. While it might be seen as just another item on the ever-growing list of business challenges, its transformative potential is a game-changer.”

AI fraud: Why you can no longer trust your senses

So what can we as consumers do? Anyone who believes that trained senses alone are enough to… Avoid AI fraudhe is wrong. A British study shows that even people with excellent facial recognition only recognize fake faces 41 percent of the time. The average person does significantly worse.

So we can’t rely on our brains – ours reliable companion in the last millennia of evolution – leave. That’s why we have to build as large a safety net as possible. Would you like an example? When making suspicious calls, you should always ask for a phone number to call back and the relevant office – this works great, for example, so that you don’t fall for AI police or fake bank employees.

This is equally important Multi-factor authentication. So: Don’t just use a password and rely on SMS codes, as these can now be intercepted. That’s why it’s better to rely on biometric keys such as your face or fingerprint. In the times of the AI ​​revolution our humanity our best protection.

Also interesting: