Tech Times
Posts
TechTimes Digest: AI Exploits, Deepfake Chaos, and One-Click Privacy

TechTimes Digest: AI Exploits, Deepfake Chaos, and One-Click Privacy

Justin Nam
October 10, 2025

Google Ignores Hidden Gemini AI Exploit That Lets Hackers Control Text

(Photo: Vincent Feuray/Getty Images)

Cybersecurity researchers discovered a severe "ASCII smuggling" exploit in Google's Gemini AI, allowing attackers to manipulate the AI’s text responses with invisible commands. Google has controversially declined to patch the flaw, labeling it a 'social engineering' issue, despite other major AI models blocking the vulnerability.

CHECK OUT →

AI-Generated Protest Videos Made With OpenAI's Sora 2 Spark Outrage, Confusion Online

(Photo: Alex Radelich/Unsplash)

Highly realistic protest videos created using OpenAI’s Sora 2 are flooding social media, raising serious concerns about misinformation and political propaganda. The deepfakes are increasingly difficult to distinguish from real footage, fueling a growing 'truth crisis' despite the presence of watermarks.

READ MORE →

California Just Made It Easier to Stop Companies From Selling Your Data With One Click

(Photo: Chip Somodevilla/Getty Images)

California Governor Gavin Newsom signed a landmark privacy law that mandates web browsers to include an easy, single-click opt-out feature. This new legislation significantly expands the California Consumer Privacy Act (CCPA), making privacy protection a default standard against the third-party sale of personal data.

READ MORE →

Until next time,

AI EXPLAINED team