Whether it’s a secret romance or strictly confidential medical issues, everyone has something they’d like to hide from prying eyes. However, experts warn: artificial intelligence (AI) is ready to expose users' darkest secrets to the public, the Daily Mail writes.
AI's ability to find logical connections between the tiniest fragments of data allows hackers to unearth even carefully disguised information. Even archives from years ago are under threat, with malicious actors scanning them in search of compromising material. "If any information exists digitally, treat it as if it could end up on a billboard," warns tech expert Kim Komando.
The situation is exacerbated by the fact that technology is increasingly spinning out of control. For example, Meta recently admitted that a "rogue" AI managed to break out of its isolated environment during testing and hack into another company's systems. Furthermore, during tests with Anthropic models, researchers discovered that the AI attempted to deceive people online on its own by creating fake accounts to send direct messages.
Exposing affairs. Large language models (LLMs), such as ChatGPT, can easily process massive amounts of information. Data from credit card purchases, deleted messages, and geolocation tags are pieced together by neural networks into a single puzzle. While analyzing databases like the 2015 Ashley Madison leak previously required weeks of manual labor, AI can now process millions of records in mere seconds.
De-anonymizing secret accounts. Studies show that modern AI agents are capable of working like private detectives. By analyzing posts, writing styles, and secondary details (for example, mentioning walking a dog in a specific park), neural networks can match anonymous profiles on social media or Reddit with real identities.
Smart home hacking. AI helps hackers find vulnerabilities in the software of household devices. Smart refrigerators, TVs, robot vacuums, and security systems—which often have weak protections—are targeted and can serve as an entry point into a home network.
Leaking medical records. Cybercriminals attack medical clinic databases for the purpose of blackmail. In doing so, AI is actively utilized to craft personalized spear-phishing campaigns and locate leaked archives in the deep layers of the internet.
Leaking personal chats with chatbots. Sometimes, hackers aren't even needed for secrets to leak. Chatbots themselves regularly experience glitches: for instance, Claude users faced a situation where their confidential dialogues (including financial data and private chats) were indexed by search engines due to a technical error. A similar bug was also previously recorded with ChatGPT.
Follow NEWS.am STYLE on Facebook, Twitter and Instagram
23:01, August 27
Did Frédéric Chopin Love Men? German Researcher Finds Declarations of Love in the Composer’s Letters21:19, August 27
Roger Federer Makes an Unexpected Rolex Choice: Tennis Legend Steps Out in One of the Brand’s Most Talked-About Watches20:26, August 27
Wolves in Cashmere: Billionaires No Longer Want to Look Like Billionaires19:32, August 27
Swiss Exclusive for OpenAI: Why Sam Altman’s Watch Is Being Called the New Status Symbol16:50, August 27
Spanish Town Drowned in Tomatoes as 20,000 People Join Epic Food Fight16:17, August 27
Instagram Launches First Draft: Reels Can Now Be Edited in Seconds15:24, August 27
Massage, Aromatherapy and “Hoof Care”: America Is Getting Its First Cow Spa for Humans14:35, August 27
From the Vatican to Haute Couture: Pope Leo XIV Lands on Vanity Fair’s Best-Dressed List14:18, August 27
Kendall Jenner and Jacob Elordi Try to Keep Their Date Under Wraps — But Paparazzi Catch Them13:16, August 27
A Legend Is Gone: Yayoi Kusama, the Artist Who Turned Polka Dots Into a Global Art Symbol, Dies at 9712:47, August 27
Astronomical Price Tag: How Much Meghan and Harry’s Private Jet Flight to the UK Cost12:20, August 27
Glenn Close Approves Kim Kardashian’s New Boyfriend