The rapid development of artificial intelligence aids criminals, for example, by making it simpler to generate a large number of convincing bait letters.
Tanel Tammet, a professor of applied artificial intelligence at Tallinn University of Technology (TalTech), told ERR that the development of artificial intelligence is unlikely to radically change the nature of cyber fraud.
"However, it is certainly growing in mass, which is due to the fact that previously, all forms of fake bills of shipment, fraudulent invoices, and similar activities required a substantial amount of manpower." Artificial intelligence such as ChatGPT can compose messages on their behalf now. This makes the production of such letters cheaper, easier, recipient-specific and even more credible.
"I would not be surprised if they start including artificial intelligence photos, such as when someone says they have been in an accident and want you to pay them money, and they attach a picture of the accident. This image is also a fake made by artificial intelligence," Tammet said.
Tarmo Hanga, chief specialist at the Information System Authority (RIA) said that criminals can use artificial intelligence to develop malware as well, although this is less of a concern now.
"The quality of AI-generated malware is lower to that of malware written by a skilled hacker," he said.
Tammet added that instead of writing their own malware, criminals are purchasing it from more experienced cyber criminals.
If you instruct an artificial intelligence to attack or clog up other people's computers, it is unlikely that it will do so successfully. When this happens, it always exhibits the same pattern. "This product type is already widely available," Tammet said.
However, despite the fact that fraudulent letters are not a new phenomenon, the number of them could rise. "We need to be vigilant," he said. According to Hanga, artificial intelligence has a generally consistent writing style; it is written professionally.
A more widespread trend is the creation of websites using computer-generated text to lure people into reading in order to generate revenue from displaying advertising. Machine-written text may mimic news stories, for example, but they would contain entirely false information.
"You can get a lot of very lofty text generated but it would not be very meaningful. So the internet space will get cluttered with this type of filler very quickly. That is something we will seeing in a few years' time."
Editor: Kristina Kersa