Three main threats of AI: Warning from an information security expert

With the development of artificial intelligence, millions of people use chatbots daily for a wide range of tasks — from writing texts to trip planning. However, not everyone is aware of the associated risks. Svetlana Gazizova, product owner for AI security at UserGate, highlighted three main threats that Russian users face in an interview with Gazeta.Ru.

1. Disinformation and blind trust in AI responses

One of the most common dangers is excessive trust in information generated by neural networks. AI can confidently produce inaccurate or outdated data because it relies on various sources without proper verification.

Gazizova cited a real-life example: a family planning a foreign vacation checked visa requirements with a neural network. The AI confidently replied that no visa was needed. Tickets were purchased and a hotel booked, but at the airport, passengers were denied boarding due to missing documents. The result was financial loss and a ruined vacation.

The expert recommends always cross-checking critically important information from official sources. For example, when asking about visas, one should specify in the prompt to use data specifically from the embassy or consulate website.

2. Confidential data leaks

The second serious threat is the transmission of personal and corporate information to chatbots. Many users upload sensitive data to AI without hesitation, not considering how it might be used further.

Gazizova emphasized that cases of leaks caused by AI's access to corporate and personal data have already been recorded, adding that under no circumstances should one upload passport scans, financial documents, or any sensitive information into chatbots.

According to her, uploaded data may be used for model fine-tuning or, in some cases, become accessible to third parties.

3. Harmful content for minors

The expert paid special attention to children's safety. Underage users can easily access information through language models that is not intended for their age: information about prohibited substances, methods of destructive behavior, and other harmful content.

Against this backdrop, regulators in various countries are tightening control over AI services and introducing additional restrictions to protect children.

Why this matters right now

According to Gazizova, tens of millions of Russians actively use AI chatbots, but basic digital hygiene when working with them remains low. Experts recommend treating AI as a useful but not always reliable assistant that requires mandatory fact-checking and caution with data.

In brief

InfoSec expert Svetlana Gazizova identified three main risks of using AI for Russians: the spread of disinformation due to blind trust, confidential data leaks, and children's access to undesirable content. The specialist warns that one should not rely on AI for important matters without cross-verification, upload personal documents to chatbots, or leave children unsupervised. Following simple digital safety rules will help avoid serious problems.