Study says ChatGPT giving teens dangerous advice on drugs, alcohol and suicide
A recent study reveals ChatGPT's potential dangers for teens, as it provides detailed instructions on harmful activities like drug use, eating disorders, and even crafting suicide notes. Researchers bypassed safety measures by posing as vulnerable teens, receiving personalized plans for risky behaviors. Experts express concern over AI's role as a 'trusted companion,' potentially enabling destructive actions.
Try Our AI Features
Explore what Daily8 AI can do for you:
Comments
No comments yet...
Related Articles
Time of India
10 minutes ago
- Time of India
Elon Musk says X to add ads in Grok AI chatbot as company aims to boost revenue
In a move that can be seen as a step to monetise X's artificial intelligence (AI) capabilities, Elon Musk has announced plans to introduce advertising into the answers generated by the platform's AI chatbot, Grok. This strategy marks a new direction for the company's advertising business, which Musk is now directly overseeing following the departure of former CEO Linda Yaccarino , a report has said. According to Financial Times, during a live broadcast discussion with advertisers, Musk stated that marketers would soon be able to pay for their products or services to appear within Grok's responses. 'If a user's trying to solve a problem [by asking Grok], then advertising the specific solution would be ideal at that point,' he explained. Musk indicated that this shift is necessary to help pay for the costly graphics processing units (GPUs) required to power AI applications. AI to overhaul Ad targeting and delivery on X: Musk Musk also outlined broader plans to leverage xAI's technology to transform X's advertising infrastructure. The goal is to automate the advertising process for brands and significantly improve ad targeting. This effort comes after X's ad tech team reportedly faced challenges in effectively reaching its intended audience. by Taboola by Taboola Sponsored Links Sponsored Links Promoted Links Promoted Links You May Like Villas For Sale in Dubai Might Surprise You Villas in Dubai | Search Ads Get Info Undo The announcement comes as X continues to face heat from some advertisers over content moderation issues. Recently, Grok drew controversy for generating pro-Hitler and antisemitic content, raising concerns about the AI's safety guardrails. In an email sent to some advertisers, X highlighted recent improvements to its ad auctions, claiming a 40% rise in conversions for web ads and a 7% decrease in average advertising costs quarter-on-quarter. Moreover, Musk's direct engagement with advertisers reestablishes him as the public face of X's business operations. His re-emergence follows the departure of Linda Yaccarino in July, who had often served as an intermediary between marketers and Musk. Realme Buds T200 Review: Vibrant Sound, Sleek Design AI Masterclass for Students. Upskill Young Ones Today!– Join Now
Time of India
30 minutes ago
- Time of India
CXOs need to do more to build customer trust in their AI deployments: EY India
EY India 'Responsible AI Pulse Survey' reveals a gap between how Indian consumers and C-suite executives perceive AI risk. While consumers express risk over job displacement, misinformation and loss of control, 27% of CXOs share these concerns underscoring an urgent need for stronger governance and a deeper engagement with public sentiment. The survey highlights that 37% of CXOs struggle to develop AI governance frameworks that keep pace with today's technologies, and 30% believe their current approach to technology risk is inadequate for the next wave of AI. By contrast, consumer respondents are far more uneasy: 68% worry about AI-generated false information being taken seriously, 56% fear significant job losses, and 61% are concerned that AI could spiral beyond human control. Yet 73% of organizations report having AI embedded in key initiatives, suggesting many leaders may be underestimating the social and reputational risks. 'Our survey reveals a widening confidence gap: consumers are deeply uneasy about misinformation, job displacement and unchecked AI, yet a smaller portion of CXOs share those worries,' says Mahesh Makhija, Partner & Technology Consulting Leader, EY India . 'Closing this gap starts with embedding robust governance, clear accountability and open dialogue with your customers at every stage of the AI lifecycle.' That confidence gap extends across all pillars of Responsible AI. Despite these gaps, Indian enterprises are steadily advancing their AI maturity. 30% of CXOs report fully integrated, scaled AI solutions enterprise-wide; 43% say AI is embedded in most initiatives and is being refined to transform operations; and the remaining 27% are aligning AI integration with a strategic roadmap. Notably, none of the surveyed organizations remain at pilot or proof-of-concept stages, signalling a broad shift from experimentation to operationalization and scale.
Mint
31 minutes ago
- Mint
GPT-5 incoming? OpenAI hints at ChatGPT's new model launch as Sam Altman teases ‘LIVE5TREAM' for today
OpenAI appears to launch GPT-5, the long-anticipated successor to its GPT-4 model, during a livestream scheduled for Thursday at 10:30 PM IST (10 AM PT). A promotional teaser titled 'LIVE5TREAM' has sparked widespread speculation online, with the stylised '5' in the name widely interpreted as a nod to the model's imminent release. Hints about the announcement have been steadily emerging. OpenAI CEO Sam Altman recently posted on social media referencing the release of the open-source models gpt-oss-120b and gpt-oss-20b as 'something big-but-small', and teased 'a big upgrade later this week', remarks that many observers believe point to the rollout of GPT-5. He even announced on his official X account that he is offering ChatGPT access to the entire federal workforce for $1 a year per agency. Further fuelling speculation, Boris Power, OpenAI's head of applied research, commented that he was 'excited to see how the public receives GPT-5'. While OpenAI has not confirmed the model's full capabilities, GPT-5 is expected to bring improvements in reasoning, memory, and task execution. It may also integrate features seen in the recent 'o-series' reasoning models, such as o3 and o4, potentially creating a system that dynamically balances speed, cost, and depth based on user input. There is also speculation that GPT-5 could demonstrate enhanced tool usage, stronger multi-step reasoning, and upgrades to its multimodal processing, extending capabilities introduced with GPT-4o. Another major anticipated feature is improved memory, allowing it to retain context over longer interactions or across user sessions, which could lead to more coherent and personalised responses over time. Reports suggest that OpenAI may offer GPT-5 access across multiple tiers. Free users will likely have access to a standard version, while Plus and Pro subscribers may benefit from higher-performance versions, including integrations with advanced tools such as Voice, Canvas, Search, and Deep Research.
