The increase in ‘Internet of things’ and popularity of Artificial intelligence or internet applications like Chat GPT and other generative AI tools has empowered us. These AI applications have diverse applications ranging from Medicine to Business. Therefore they also do come with hidden risks.
Some of the risks or concerns associated with Generative AI are explained below:
- Cybercrime
Europol, the EU Agency for Law Enforcement, has expressed its concern that these AI tools have the potential to increase Cybercrime. The ease with which convincing contents are generated through different AI tools can be exploited for illegal activities.
For Example,
- AI generated Phishing Emails
Phishing emails are generated and with the help of AI, they mimic the legitimate email, making it more believable.
- AI impersonation
Cybercriminals carry out scams through Vishing. Vishing is a type of scam which occurs through a phone call. Here the cybercriminals target the victims and pretend to be friends, colleagues or family members of the victim.
- Deep Fakes
Deep Fakes are fake forms of media created using AI. They can spread misinformation or malinformation.
- Spread of Misinformation
Cybercriminala can create Deep Fakes and spread misinformation. With the help of AI, a person’s face and body are altered to make them look like a person they are not. This can have serious implications during National elections and lay a profound impact on democracy.
- Black-box nature: unreliability & interpretability issues
Many AI systems operate as black boxes. It is hard to understand the whereabouts of their predictions and responses making them unreliable with the lack of transparency and interpretability.
- False Accusation
Large Language Models like ChatGPT are known to hallucinate. They can provide false statements or generate results that are not accurate. For example, ChatGPT had falsely accused a Law professor of sexual harassment. This happened because ChatGPT was confused by irrelevant connections between different pieces of texts.
AI tools come with incredible potentials as well as significant challenges. That is why there is a dire need to regulate AI systems through an International Agency.