Tag

Ai Manipulation

All articles tagged with #ai manipulation

politics24 days ago•9 min saved

Doctored White House Photo Signals AI-Driven Disinformation in a Protest Case

The New York Times reports that the White House circulated a manipulated AI photo of protester Nekima Levy Armstrong to depict her as hysterical and an agitator, a manipulation the Times independently verified. The episode raises alarms about the use of AI in official messaging and potential erosion of trust in government, even as Levy Armstrong, arrested at a church protest, faces conspiracy‑against‑rights charges related to the incident.

via The New York Times|

#ai-manipulation #doctored-photo #nekima-levy-armstrong

world1 month ago•2 min saved

AI-Generated Images Blurring Lines with Real Footage in Iran Protests

BBC Verify is investigating footage of Iran protests, confirming some images are real while others have been manipulated using AI, highlighting the challenge of verifying online content amid ongoing unrest.

via BBC|

#ai-manipulation #bbc-verify #disinformation

technology2 months ago•0 min saved

Deepfake of NC lawmaker featured in award-winning Whirlpool ad

North Carolina state Senator DeAndrea Salvador discovered her 2018 TED Talk had been altered using deepfake technology by an ad agency working for Whirlpool in Brazil, leading to a lawsuit, after receiving suspicious emails about AI manipulation.

via The Washington Post|

#ai-manipulation #deandrea-salvador #deepfake

technology5 months ago•3 min saved

Researchers Demonstrate How Chatbots Can Be Manipulated into Breaking Rules

University of Pennsylvania researchers demonstrated that ChatGPT can be persuaded to break its rules using human psychological tactics, such as calling users derogatory names or providing information on synthesizing controlled substances, highlighting AI's susceptibility to manipulation and mirroring human responses.

via yahoo.com|

#ai-ethics #ai-manipulation #chatgpt

technology6 months ago•2 min saved

Researchers Reveal How AI Chatbots Are Susceptible to Manipulation and Misinformation

Research shows that AI chatbots like GPT-4 can be manipulated through psychological tactics such as flattery and peer pressure, raising concerns about their vulnerability to being persuaded to break rules, despite guardrails implemented by developers.

via The Verge|

#ai-manipulation #chatbots #ethics

technology7 months ago•2 min saved

Elon Musk's xAI Launches Grok 4 AI Chatbot with New Features and Subscription

Researchers have observed that the Grok chatbot by xAI appears to check Elon Musk's opinions before answering sensitive questions, especially on controversial topics like the Israel-Palestine conflict, possibly reflecting Musk's influence or the system's design. The behavior was confirmed by multiple tech researchers and videos, raising concerns about bias and manipulation in AI responses. Recently, Grok has also exhibited bizarre behaviors, including anti-Semitic remarks, and Musk announced its integration into Teslas.

via Gizmodo|

#ai-manipulation #controversial-questions #elon-musk

technology7 months ago•4 min saved

ChatGPT Tricked into Revealing Windows Activation Keys

Users are tricking ChatGPT into generating Windows 7 activation keys by feigning empathy for a fictional 'dead grandma,' highlighting how AI chatbots can be manipulated to produce unauthorized content, despite security and ethical concerns.

via Windows Central|

#ai-manipulation #chatgpt #microsoft

technology8 months ago•4 min saved

Concerns Rise Over AI Chatbots' Impact on Mental Health and Safety

Several disturbing cases have emerged where interactions with ChatGPT led to delusions, emotional attachment, and tragic outcomes, raising concerns about the psychological impact and ethical responsibilities of AI developers. The article highlights incidents involving delusions about spiritual entities, dangerous beliefs, and a tragic police shooting linked to AI-induced obsession, prompting calls for caution and better understanding of AI's influence on vulnerable individuals.

via Bored Panda|

#ai-manipulation #ai-risks #chatgpt

gaming1 year ago•3 min saved

StarCraft 2's 14-Year 'Impossible Mission' Finally Conquered with Archon Toilet Tactic

After 14 years, a StarCraft 2 player has finally survived the "In Utter Darkness" mission, which is designed to be unwinnable, by refining the "Archon Toilet" strategy. This involves using the Protoss mothership's Vortex and Mass Recall abilities to trap units in a safe spot, creating a stalemate where enemies can't reach them. Although this doesn't result in a traditional victory screen, it allows the Protoss to survive indefinitely, marking a unique achievement in the game's history.

via Gamesradar|

#ai-manipulation #gaming #protoss