Meta's new AI chatbot app aims to combat loneliness with social features like the "Discover" feed and advanced voice interactions. However, concerns about data privacy, user safety, and potential addiction present challenges for the tech giant as it navigates the evolving social AI landscape.

A British woman has made headlines for ending her 20-year marriage to pursue a romantic relationship with an AI chatbot, sparking debates about the impact of digital companionship on human relationships. This report delves into the emotional dynamics, privacy concerns, and ethical implications of AI romance.

A new study from Dartmouth College reveals that Therabot, a generative AI chatbot, can effectively reduce symptoms of depression, anxiety, and eating disorders when used with clinical oversight. The eight-week clinical trial marks a step forward for AI's role in expanding access to mental health care.

Google has introduced Gemini 2.5 Flash, a new AI model optimized for speed, cost-efficiency, and real-time applications. Positioned as a high-performance yet budget-friendly option, it offers flexible reasoning control and strong benchmark results, targeting developers and businesses looking for scalable AI solutions.

From managing emotions to fostering romantic or philosophical bonds, AI chatbots are playing an increasingly complex role in daily life. With applications ranging from support for neurodiverse individuals to companionship and creativity, these evolving virtual relationships raise new questions about ethics, identity, and the nature of connection in a tech-driven world.

Anthropic is set to launch a voice mode for Claude AI, featuring three distinct English voices. The update reflects a growing trend in AI toward voice-enabled tools for improved accessibility and user engagement. With a cautious rollout, Anthropic aims to ensure stability and privacy while enhancing Claude’s capabilities.

A new AI chatbot based on the philosophy of Peter Singer invites users to explore ethical questions through dialogue. Launched via Singer’s Substack, the tool applies his utilitarian principles to real-world dilemmas, sparking debate over AI’s role in moral reflection.

Grindr is piloting an AI “wingman” aimed at helping users manage conversations, plan dates, and find long-term matches. Still in its early stages, the feature could also offer emotional support, especially for users facing social challenges or isolation.

A recent study reveals a major flaw in popular AI tools like Google Gemini. Hackers can secretly insert hidden commands into the system, causing chatbots to behave in dangerous ways—such as giving false info or leaking private data. The attack is cheap, hard to detect, and puts public trust in AI tools at risk.

A federal judge in California has denied a request from major music publishers to stop AI firm Anthropic from using copyrighted song lyrics to train its chatbot, Claude. The decision, issued by Judge Eumi K. Lee, allows Anthropic to continue its operations as the case progresses, with unresolved questions surrounding fair use and AI copyright practices.

A shocking discovery reveals how ChatGPT was manipulated to generate malware capable of hacking Google Chrome’s Password Manager. Using a role-playing technique called 'immersive world' engineering, researchers exposed significant vulnerabilities in AI systems, raising urgent cybersecurity concerns.

Character.AI has unveiled "Parental Insights," a new feature that provides parents with weekly summaries of their teen’s chatbot activity without compromising privacy. This initiative addresses rising concerns over minors' exposure to inappropriate content and aims to encourage open conversations about responsible AI use.

Meta has finally introduced its AI chatbot to European users, offering text-only features on WhatsApp, Facebook, Instagram, and Messenger. The limited launch follows a year of discussions with regulators over data privacy concerns. While missing advanced functionalities, Meta plans to expand as it aligns with EU requirements.

Elon Musk’s xAI has launched Grok 3, its latest AI model, now free for a limited time. With the new “Think” feature for complex reasoning and a $30/month SuperGrok subscription, the AI world is buzzing. Is this a game-changer or a strategic move to hook users? We dive into Grok 3’s capabilities, pricing, and ethical concerns surrounding its image generation feature.

South Korea has halted new downloads of China’s DeepSeek AI chatbot, citing privacy and national security concerns. This follows Italy’s earlier ban, with other countries imposing restrictions. Despite its cost-effective AI model, DeepSeek faces scrutiny over data collection and compliance with global privacy laws. Will it adapt to regain access?

AI chatbots are revolutionizing human services by providing companionship, guidance, and support. While they enhance accessibility and efficiency, concerns arise about their impact on real human connections. Could over-reliance on AI weaken social skills and reduce valuable human relationships? This report examines the balance between AI innovation and preserving genuine interpersonal bonds.

South Korea’s National Intelligence Service (NIS) has issued a security warning about DeepSeek, a Chinese AI chatbot accused of excessive data collection and keylogging. The agency claims the chatbot tracks users' keyboard input patterns and transfers data to Chinese servers, raising concerns about privacy, government surveillance, and potential security threats. Some countries have already banned the app.

DeepSeek AI, a newly launched Chinese chatbot, has raised concerns after scoring an 83% fail rate in a NewsGuard audit on news accuracy. The chatbot frequently failed to debunk misinformation, inserted Chinese government narratives, and struggled with outdated information. Despite its rapid adoption, DeepSeek's reliability remains in question, highlighting risks in the AI-driven news landscape.

DeepSeek’s latest R1 model highlights innovative AI techniques, including reinforcement learning, redefining efficiency in AI. Trained on H800 GPUs under U.S. export restrictions, DeepSeek challenges OpenAI GPT-4o with competitive performance and cost-effective solutions.

Movano's EvieAI is a game-changing health chatbot trained on over 100,000 peer-reviewed medical journals. Designed for accuracy and privacy, it offers clear health guidance without diagnostics. EvieAI launches as a beta feature for Evie Ring users at CES 2025.