AI That Knows Your Personal Details May Prompt You to Click a Link

Artificial intelligence is raising new concerns about online privacy and cybersecurity as increasingly personalised systems create opportunities for more sophisticated phishing attacks. Security experts warn that personal information, when combined with AI-generated messages and deceptive links, could make it harder for users to distinguish genuine communications from scams.

AI assistants are becoming more capable of understanding user preferences, processing information and helping complete online tasks. While these features offer greater convenience, they also raise questions about how personal data could be misused and whether people might place too much trust in messages that appear tailored specifically to them.

Traditional phishing scams often rely on suspicious emails, fake warnings or generic requests for passwords and financial details. AI tools, however, can help generate messages that sound natural and include relevant personal or professional information, potentially making fraudulent communications more believable. A message referring to a recent purchase, a workplace project or an upcoming appointment may appear legitimate at first glance. If it contains a link requesting account verification or access to a document, recipients could be more likely to click without checking its authenticity.

The growing use of AI agents introduces another layer of concern. Depending on their capabilities and permissions, these systems can browse websites, retrieve information and interact with digital services on behalf of users. Security researchers have warned that malicious instructions hidden in online content could attempt to manipulate such systems into taking unintended actions or exposing sensitive information.

Privacy is also becoming a central issue as digital services collect information from multiple sources. Details shared through social media, online accounts and workplace platforms may collectively reveal more about an individual than any single piece of information. However, the information available to an AI system varies according to the product, user permissions and connected services.

Cybersecurity specialists recommend that users independently verify unexpected requests involving payments, passwords, identity documents or account permissions. Rather than clicking a link in a suspicious message, users should visit the relevant service through its official website or application. Reviewing application permissions, enabling multi-factor authentication and using unique passwords can also help reduce the risk of account compromise. Users should be particularly cautious when an unfamiliar application requests access to emails, contacts, files or other sensitive information.

Technology companies developing AI systems face the additional responsibility of protecting private data, limiting unauthorised access and ensuring that automated systems handle untrusted online content safely. As artificial intelligence becomes more integrated into everyday digital activities, the ability to personalise information will remain one of its most useful features. However, the same capability could also be exploited to make online fraud more convincing. The emerging challenge for users and technology companies is to ensure that greater personalisation does not come at the expense of privacy, security and informed decision-making.

Disclaimer: This image is taken from Reuters.