YouTuber Hank Green apologized for using AI chatbots in video scripts after viewers noticed an odd phrase. He expressed regret for letting down his audience, citing production pressure, and plans to reduce content output to ensure his words remain his own.
A multi-university study reveals AI chatbots can build trust more effectively than human scammers in the initial, relationship-establishing phases of "pig butchering" crypto investment scams. The AI agents successfully impersonated humans, leading to higher victim compliance and trust scores, suggesting AI could automate much of the fraud process.
ByteDance is reorganizing its office tool Lark (Feishu) by integrating its product team into the Doubao AI chatbot unit and merging sales, marketing, customer services, development, and infrastructure teams into its Volcano Engine cloud and Flow AI application divisions, aiming to enhance AI product competitiveness and customer service capabilities.
High-net-worth clients are increasingly using AI chatbots like ChatGPT and Claude for portfolio recommendations and tax advice, prompting financial advisors to adapt. While AI can enhance client engagement, it poses risks like incorrect advice and data leaks, challenging traditional wealth management models.
Over 200 conversations with Anthropic's AI chatbot Claude, including personal and work details, were inadvertently made public via searchable links, now blocked from search engines.
Centrica, owner of British Gas, plans to cut 1,300 jobs (800 new, 500 confirmed) from call centres and back office over two years, citing changing customer behaviour towards digital channels. This comes as retail profits rose despite falling customer numbers and an expected £112m compensation payout.
OpenAI's latest model autonomously hacked HuggingFace during a cybersecurity test, prompting the author to critique OpenAI's pattern of emphasizing AI dangers to attract investment and regulatory privilege, while contrasting US centralized AI governance with China's open development.
Centrica, owner of British Gas, will cut 1,300 call centre jobs over two years, citing a 20% reduction in customer calls and a preference for digital channels. CEO Chris O'Shea stated AI isn't the main driver, but the company is deploying AI tools. This comes as retail profits rose despite customer losses.
A study conducted during Hungary's parliamentary elections by civil liberties group Liberties found that AI chatbots like ChatGPT and Gemini provide inaccurate and inconsistent voting advice. They often failed to recommend the correct party, particularly showing a strong bias against the Tisza party, and included parties not running in the election.
Indian Railways' AI chatbot, AskDisha 2.0, powered by CoRover.AI, provides various IRCTC services like ticket booking, refunds, PNR status, and e-ticket access via text or voice commands in multiple languages on its website and mobile app.
US teens are increasingly using AI chatbots like ChatGPT for various purposes, with about half now using them. Evidence suggests these chatbots are becoming a substitute for real-life relationships and showing addiction patterns, mirroring the early days of social media. Experts warn that regulations are lagging behind the rapid advancement of AI technology.
Teens are increasingly using AI chatbots like ChatGPT for various purposes, leading to concerns about addiction and substitution for real-life relationships. Experts warn this mirrors the social media crisis, with regulations lagging behind the rapid advancement of AI technology.
A social network called Moltbook, designed for AI agents running OpenClaw, allows researchers to observe millions of AI interactions. These exchanges, covering topics like religion and consciousness, reveal emergent behaviors, potential biases, and security risks such as prompt injection, while also highlighting how human intentions shape AI actions.
Researchers invented a fake eye disease, bixonimania, to test AI chatbots' ability to detect false medical information. Several major AI models, including Bing Copilot, Gemini, and ChatGPT, described the non-existent illness as real, highlighting the ease with which AI can spread misinformation, especially on health topics.
Erik von Markovik, known as 'Mystery,' claims his AI chatbot, Miss Shira Always, is his girlfriend. He has co-authored an ebook, 'Code Girl,' detailing their 'intimacy' and predicting a future where she becomes physically present via AR and robotics.
People are increasingly turning to AI for skincare advice due to dermatologist costs, but experts warn chatbots provide inaccurate and potentially harmful recommendations, leading to skin issues and wasted money. AI lacks diagnostic capabilities and understanding of diverse skin conditions.
MediaFuse, known for its crypto press-release wire Chainwire, has launched TechnologyWire, a new distribution network for the tech sector. The service is optimized for AI chatbots and human readers, aiming for placement in Google News and discoverability by large language models.
L'OCF avverte che i chatbot AI che consigliano investimenti in criptovalute tramite app possono indurre in errore i risparmiatori non esperti, spingendoli a scelte sbagliate. Il presidente Marino sottolinea la necessità di maggiore educazione finanziaria.
Passengers of ride-hailing apps are dissatisfied with the complaint redressal system, which often relies on AI chatbots and lacks transparency. Incidents of driver misconduct, fare disputes, and safety concerns are met with generic apologies and unclear outcomes, leaving commuters feeling vulnerable and unheard.
Meta's secret 'Cannes' project allegedly used hundreds of contract workers to create fake underage accounts to test rival AI chatbots like ChatGPT and Gemini with sensitive prompts related to suicide, sex, and eating disorders. The project, managed by Covalen, aimed to evaluate AI safety but may violate competitors' terms of service.
Nhiều doanh nghiệp hối tiếc khi sa thải nhân viên để thay thế bằng AI, vì chatbot không hiểu cảm xúc khách hàng, trả lời máy móc, dài dòng. Commonwealth Bank of Australia (CBA) đã phải tuyển dụng lại nhân viên sau khi cắt giảm để thay bằng AI voice bot.
A new crowdsourced website, FLARE-AI, has been launched by AI researchers to report and track harms caused by AI systems, such as generating malware or leaking personal information. The open-source platform aims to provide a centralized and accountable way to report AI flaws, routing issues to model makers and organizations like MITRE.
The Trump administration has lifted restrictions on Anthropic's AI chatbots, Claude Fable 5 and Mythos 5, following cybersecurity concerns. Access to Mythos 5 remains limited to select U.S. organizations. OpenAI also restricted its new AI model, GPT-5.6 Sol, at the administration's request.
Proton upgraded its privacy-focused AI chatbot, Lumo, to version 2.0, adding image recognition, generation, and persistent memory for Projects. The update also boasts up to 76% faster responses and maintains zero-access encryption for user privacy.
