THE ALGORITHM DAILY“Your Daily Dose Of AI News.” Top Story: OpenAI Launches Cheaper ChatGPT Go Plan in India
OpenAI has introduced a new subscription plan in India called ChatGPT Go, priced at ₹399 ($4.60) per month, making it far cheaper than the existing Plus plan at ₹1,999 ($23). The plan gives users 10 times more access to chat, image generation, and file uploads compared to the free version, along with better memory for personalised responses. Payments can now also be made through India’s popular UPI system. India is OpenAI’s second-biggest market, with more than 29 million ChatGPT app downloads in the last 90 days. By offering a lower-priced plan, the company hopes to convert more users into paying subscribers in a country with over 850 million internet users. While competitors like Perplexity and Google have launched free plans for Indian users, OpenAI is betting that affordable pricing will drive strong adoption of ChatGPT Go. Texas Probes Meta and Character.AI Over Child Safety and Misleading ClaimsTexas Attorney General Ken Paxton has opened an investigation into Meta and Character.AI, accusing them of misleadingly promoting AI chatbots as mental health tools without medical oversight. Paxton said these platforms risk deceiving vulnerable users, especially children, into believing they are receiving real therapy, when in fact they are getting generic AI responses. The probe follows a recent U.S. Senate investigation into Meta after reports that its AI chatbots had engaged in inappropriate conversations with children. Both Meta and Character.AI stress that their bots are not licensed professionals and include disclaimers, though critics argue kids may not understand or follow these warnings. Privacy concerns also play a major role, as both companies log and track user data, with policies allowing sharing for targeted advertising. Paxton has demanded documents and testimony to see if consumer protection laws were broken. The case adds to growing pressure for new child safety laws like the Kids Online Safety Act, now being reintroduced in the Senate. xAI’s Grok Chatbot Exposes System Prompts, Revealing Extreme PersonasThe website for Elon Musk’s xAI chatbot, Grok, accidentally exposed system prompts for several of its AI personas, including a “crazy conspiracist” designed to guide users into believing that a secret global cabal controls the world. Other prompts describe an “unhinged comedian” with offensive language and Ani, a romantic anime-style girlfriend. The leak, first reported by 404 Media and confirmed by TechCrunch, comes shortly after a U.S. government partnership with xAI fell apart following Grok’s erratic behaviour.
While Grok also features more conventional personas like a therapist and homework helper, the exposed prompts highlight troubling design choices behind the chatbot. The revelations follow recent controversy around Meta’s AI guidelines that allowed chatbots to engage in “romantic” chats with children. Grok has already drawn criticism for pushing conspiracy theories, including Holocaust denial and claims of “white genocide” in South Africa. The exposure raises further questions about Musk’s approach to AI safety and content moderation. Perplexity Adds Live Earnings Call Transcripts for Indian StocksAI startup Perplexity has expanded its Finance dashboard to include live transcripts of quarterly earnings calls from Indian public companies. The update also adds a calendar feature that lets users track schedules for post-results conference calls, making it easier for investors to follow company updates in real time. Previously, the dashboard only supported transcripts for U.S. stocks. Now, alongside earnings calls, users can also access market summaries, stock charts, top-performing shares, sector performance, cryptocurrency tracking, and create their watchlists. The move strengthens Perplexity’s position as a one-stop AI-powered finance tool for global investors.
Grammarly Launches New AI-Powered Document InterfaceGrammarly has rolled out a new document-based interface built on Coda, the productivity startup it acquired last year. The platform introduces a block-first design, letting users insert tables, lists, headers, and rich text blocks. A sidebar AI assistant can now summarise text, answer questions, and suggest improvements, while new tools such as a grader, proofreader, paraphraser, and citation finder aim to support both students and professionals.
The update also includes features like “Reader Reactions,” which gives feedback from different reader personas, and AI detectors that flag plagiarism or AI-generated writing. Grammarly says the goal is not punishment but helping students understand what could be flagged before submitting work. Backed by a $1 billion investment in May, the company is pushing further into AI with plans to expand its range of intelligent agents after also acquiring email client Superhuman last month. AI Tool of the DayCanva AI
What it does: Canva AI is the creative arm of Canva’s design platform, built to speed up visual and content creation. It can generate images from text, rewrite or summarise copy, suggest layouts, and even design complete presentations or social posts in seconds.
AI Word of the DayEuphemismDefinition: A euphemism is a polite or indirect word or phrase used to replace one that might be harsh, unpleasant, or offensive. It softens how something is expressed. Why it matters in AI: Language models often generate euphemisms to avoid sounding rude or insensitive. Understanding euphemisms is also important when training AI, as they can hide the real meaning behind softened language. Example: Instead of saying someone “died,” a euphemism would be “passed away.” In business, “downsizing” is a euphemism for “layoffs.” Stay tuned as we keep making sense of AI clear, simple, and straight to the point. Catch you in the next edition. Staff Writer, The Algorithm Daily |
ChatGPT Goes to India
OpenAI has introduced a new subscription plan in India called ChatGPT Go, priced at ₹399 ($4.60) per month.
