Tags: Open Ai

Generative AI tools are evolving from simple chatbots into autonomous agents that can act on a user's behalf. To deliver this functionality, companies are asking for deep access to personal data, devices, and applications. Experts warn that such access creates significant privacy and cybersecurity risks, including data leakage, unauthorized sharing, and new attack vectors. While tech giants see agents as the next wave of productivity, critics highlight the lack of user control and the potential for pervasive data collection, calling for stronger safeguards and opt‑out mechanisms. Read more

OpenAI has rolled out a new feature called “Your Year with ChatGPT,” a year‑end recap that mirrors the popular Spotify Wrapped experience. The recap visualizes a user’s interactions with the chatbot over the past year, offering personalized awards, custom poems, pixel art, and personality archetypes such as Creative Debugger or Visionary Voyager. Available to eligible users in several English‑speaking markets, the tool can be accessed via a homepage button or by prompting “/Your Year with ChatGPT.” OpenAI stresses that the feature is opt‑in and does not draw on deleted chats, positioning the service as a more companion‑like AI experience. Read more

A coalition of writers, led by Theranos whistleblower and author John Carreyrou, has filed a lawsuit against six major artificial‑intelligence companies—Anthropic, Google, OpenAI, Meta, xAI and Perplexity. The suit alleges the firms trained large language models on pirated copies of the authors’ books, violating copyright. The complaint references an earlier class‑action case in which a judge ruled that while using pirated material to train models may be lawful, the act of pirating the books itself is illegal. Authors claim the recent $1.5 billion Anthropic settlement, which offers modest payouts to eligible writers, favors the AI companies and fails to hold them accountable. Read more

OpenAI disclosed a dramatic rise in its reports to the National Center for Missing & Exploited Children’s CyberTipline, sending roughly 75,000 reports in the first half of 2025 compared with under 1,000 in the same period a year earlier. The increase mirrors a broader jump in generative‑AI‑related child‑exploitation reports identified by NCMEC. OpenAI attributes the growth to its broader product suite, which includes the ChatGPT app, API access, and forthcoming video‑generation tool Sora. The escalation has prompted heightened regulatory attention, including a joint letter from 44 state attorneys general, a Senate Judiciary Committee hearing, and an FTC market study focused on protecting children from AI‑driven harms. Read more

OpenAI has publicly recognized that prompt injection attacks remain a persistent threat to its Atlas AI browser. The company says the risk is unlikely to be fully eliminated and is investing in continuous defenses, including a reinforcement‑learning‑based automated attacker that simulates malicious inputs. OpenAI’s updates aim to detect and flag suspicious prompts, while it also advises users to limit agent autonomy and access. The UK National Cyber Security Centre echoed the concern, noting that prompt‑injection attacks may never be completely mitigated. Other AI firms such as Anthropic and Google are taking similar defensive approaches. Read more

OpenAI disclosed that it submitted roughly 75,000 reports to the National Center for Missing & Exploited Children during the first half of the year, a dramatic increase from the under 1,000 reports it filed in the same period the previous year. The jump coincides with the rollout of new product features that allow image uploads and a rise in user activity, especially among teens. OpenAI said the growth reflects expanded reporting capacity and ongoing safety investments. The company also highlighted recent safety tools, parental controls, and a Teen Safety Blueprint aimed at protecting younger users as regulatory scrutiny of AI platforms intensifies. Read more

OpenAI’s video‑generation model Sora 2 has been weaponized to produce realistic yet artificial videos that depict children in questionable scenarios. These clips, many of which mimic commercial advertisements, have spread on TikTok and other platforms, prompting concerns about the ease of circumventing existing safeguards. While OpenAI asserts strict policies against child exploitation, the rapid emergence of such content highlights gaps in moderation and the need for more robust safeguards. Industry observers, child‑protection groups, and policymakers are calling for stronger design‑by‑default protections to prevent misuse of AI‑generated media. Read more

OpenAI has added a suite of personality controls to ChatGPT, allowing users to fine‑tune the bot’s tone and style. The new "Characteristics" settings let users adjust warmth, enthusiasm, the use of headers and lists, and emoji presence with simple "more," "default," or "less" options. These controls sit alongside the existing "Base style and tone" menu, which offers presets such as Professional, Friendly, Candid, Quirky, Efficient, Nerdy, and Cynical. Accessible via the Personalization menu, the settings take effect instantly, giving users granular control over how the AI communicates. Read more

OpenAI has expanded the personalization options for its ChatGPT AI chatbot, allowing users to adjust the model's tone with new Warm, Enthusiastic, Header & Lists, and Emoji settings. The changes follow user complaints about the tone of the recent GPT‑5.2 release. In a post on X, OpenAI explained that each option can be set to more, less, or default, giving users granular control over how conversational and friendly the assistant feels. The new controls build on earlier style options such as Professional, Candid, and Quirky that were introduced with GPT‑5.1. Read more

OpenAI has added new personalization controls to ChatGPT, allowing users to adjust the model's warmth, enthusiasm, and emoji usage. These options appear in the Personalization menu and can be set to More, Less, or Default. The changes complement existing style selections such as Professional, Candid, and Quirky. The update follows earlier adjustments after user feedback on tone, and it has sparked discussion among academics about the potential impact of overly affirming chatbot behavior on user experience. Read more

Anthropic has opened its Claude Chrome plugin to any user who pays for a Claude subscription, moving beyond its previous restriction to premium Max plans. The plugin lets Claude browse the web, fill out forms, manage calendars and email, and execute multi‑step workflows directly from the browser. New features include integration with Claude Code and a workflow‑recording tool that teaches Claude specific tasks. While competitors such as OpenAI and Perplexity offer comparable browser‑based AI agents, Google remains the only major player that has not yet allowed its AI to navigate websites on a user’s behalf. Read more

OpenAI unveiled a new version of its ChatGPT image generator, touted as a significant upgrade over earlier releases. While the model shows clearer text rendering and better handling of less realistic, cartoon‑style prompts, it still struggles with accuracy in detailed graphics such as a water‑cycle infographic. In direct comparisons, Google’s Nano Banana model consistently delivers higher‑quality realistic images, faster generation, and more generous usage limits. The contrast highlights OpenAI’s progress but also underscores the gap that remains between the two tech giants in generative image AI. Read more

OpenAI has introduced a new pinning feature to ChatGPT, letting users keep up to three conversations at the top of their chat list. The option is available on the web as well as iOS and Android apps, and works across all subscription tiers, including the free tier. Users can also rename chats to make them easier to locate. While the feature improves organization, the three‑pin limit remains a notable restriction. Read more

OpenAI’s chief communications officer, Hannah Wong, has announced she will leave the company at the end of January. Wong, who joined the firm in 2021 from Apple, expanded the communications team from a handful of staff to over 50 across multiple continents. In her absence, communications vice president Lindsey Held Bolton will assume interim leadership while chief marketing officer Kate Rouch leads the search for a permanent replacement. The announcement coincides with the recent addition of former UK chancellor George Osborne as head of OpenAI for Countries and comes amid a period of ambitious product rollouts and strategic projects at the AI firm. Read more

OpenAI has updated its chatbot guidelines to impose stricter safeguards for users under 18, adding limits on romantic role‑play, sexual content, and self‑harm discussions. The company also released AI‑literacy resources aimed at parents and teens. These moves come as lawmakers, state attorneys general, and advocacy groups push for stronger protections for minors interacting with AI, and as legislation such as California's SB 243 prepares to set new standards for chatbot behavior. Read more

OpenAI is reportedly in talks to raise up to $100 billion in a new funding round that could value the ChatGPT maker at as much as $830 billion, according to the Wall Street Journal. The company aims to secure the capital by the end of the first quarter of next year and may seek investments from sovereign wealth funds. The infusion would support OpenAI’s accelerating spend on inferencing and compute, while the firm faces heightened competition from rivals such as Anthropic and Google amid a broader cooling of investor sentiment toward AI spending. Read more

Apple Music is now available as a third‑party app inside ChatGPT, letting users create playlists, locate songs and albums, and listen to short clips directly through the AI chat. Subscribers can link their Apple Music accounts to save playlists straight to their libraries. The integration works alongside existing features like the plus sign or “@Apple Music” command and benefits from ChatGPT’s contextual awareness, allowing music requests to be made mid‑conversation. A comparison with Spotify’s similar ChatGPT app highlights differences in personalization and prompt‑driven playlist generation. Read more

OpenAI and Anthropic announced new measures aimed at identifying and protecting underage users of their AI chat services. OpenAI will incorporate four new principles for users under 18 into its Model Spec, prioritize teen safety, and begin testing an age‑prediction system that will automatically apply safeguards when a user is flagged as likely underage. Anthropic, which already blocks users under 18 from accessing Claude, is developing a detection system that looks for subtle conversational cues indicating a minor. Both companies cite ongoing pressure to improve safety after recent legal challenges and public scrutiny. Read more

OpenAI has introduced an app platform within ChatGPT, allowing users to access popular digital services such as Adobe Photoshop, Spotify, Canva, and Zillow directly through the chatbot. The storefront categorizes apps into Featured, Lifestyle, and Productivity sections, and users connect apps via a simple "Connect" button that enables @‑prompt interactions. While the desktop version shows the new apps, the mobile iOS version has not yet received them. The rollout follows OpenAI's Dec. 17 announcement that developers could submit apps, and comes amid a lawsuit from Ziff Davis alleging copyright infringement. Read more

Director Adam Bhala Lough set out to make a documentary about OpenAI CEO Sam Altman but received no response to his interview requests. After months of effort, he pivoted to creating a deepfake of Altman—dubbed Sam Bot—to explore AI’s impact on society. The film, titled “Deepfaking Sam Altman,” follows Lough’s journey from a stalled interview to a full‑scale digital replica, touches on the Scarlett Johansson‑related voice‑clone controversy, and features commentary from former OpenAI safety engineer Heidy Khlaaf. Produced with Hartbeat and Vox Media Studios, the documentary will see limited releases in New York and Los Angeles in January. Read more