Most Americans are failing to recognize AI risks when using it at work

People in the US “know nothing about privacy issues when using AI for work,” according to a recent study by NordVPN. More often than not, they also tend to trust apps, service providers, or default WiFi settings.

The vast majority of Americans don’t think about dangers like data breaches, data misuse, bias in algorithms, lack of transparency, and even sensitive data being used for model training when using AI at work. In fact, only 5% of Americans were knowledgeable about the possible privacy issues surrounding their AI tools.

“Technological advancements might be overwhelming for some people, which may make them struggle to keep up. Convenience-driven online applications become more ingrained in our daily lives, and people often prioritize ease of use over privacy, unintentionally exposing themselves to risks,” says Marijus Briedis, chief technology officer (CTO) at NordVPN.

Generative AI technology has enhanced many internet activities, including work. However, experts warn that AI-powered applications gather large amounts of private information and may become a significant attack vector.

The research measured overall cybersecurity and internet privacy knowledge, and Americans are still among the top three countries in the world, only bested by Singapore in first place, and Finland and Lithuania tied in second.

Americans were very good at recognizing suspicious streaming service offers, creating strong passwords, avoiding malware infections, sharing sensitive data on social media, and recognizing which permissions to give to different apps. Their answers to these questions were correct 90% of the time and more.

They also know how devices might get infected with malware (93%) and which permissions to give to different apps (91%), or what sensitive data to avoid sharing on social media (90%). Americans were quite good at recognizing phishing attacks and scams.

The study also revealed that Americans trust their internet service providers (ISPs) and home WiFi networks too much. Only 15% of respondents in the US know that ISPs collect metadata about their internet usage. 16% of them consider WiFi routers to be safe with default settings. Only a quarter (27%) of Americans correctly named online tools for digital privacy protection, according to NordVPN.

The study, which gathered 25,567 responses from 181 countries, reveals that global online privacy and security awareness has been in decline over the last few years.

“It's concerning to see a decline in global internet privacy awareness, this trend only confirms the growing need for more education in protecting our personal data online,” says Briedis.

To better protect online privacy and security, NordVPN recommends creating unique and strong passwords, enabling multi-factor authentication, keeping software up to date, using a virtual private network (VPN) to encrypt traffic, and reviewing and adjusting privacy settings on social media and other apps and services. Continuous learning about cybersecurity helps to make the best-informed decisions. To take the National Privacy Test yourself, follow this link.