Latest news with #ChatGPTPro
Mint
10 hours ago
- Mint
OpenAI's GPT-5 is here — 5 game-changing features you should try
After weeks of teasing its latest artificial intelligence model to power ChatGPT, OpenAI finally unveiled GPT-5 during a live-stream on Thursday. The new AI model is claimed to outperform previous AI models from OpenAI and rival labs on various benchmarks. It is said to be better at real-world tasks, improve the ability to follow instructions, and minimize sycophancy (being overly agreeable with users). OpenAI claims improvements across the board in GPT-5 compared to GPT-4o, including accuracy, speed, reasoning, context recognition, structured thinking, and problem-solving. The biggest change is a unified system that powers GPT-5's normal tasks while a dedicated reasoning model called GPT-5 Thinking handles harder tasks. Unlike in the past, where users had to choose the model for each query, GPT-5 features a real-time router trained on real signals to instantly decide which model to use. GPT-5 is starting to roll out today and will be available for free to all ChatGPT users. However, to get higher usage volume, users will need to upgrade to a Plus subscription. ChatGPT Pro users will have access to GPT-5 Pro. OpenAI says GPT-5 makes huge strides in coding, benefiting not just software developers but also everyday users. People can now create websites, apps, and games — complete with an eye for aesthetic sensibility — using just one prompt. In its blog post, OpenAI showcased several games and apps built from simple natural language instructions. ChatGPT can now connect to Google Calendar and Gmail, letting users ask questions based on their schedule or recent emails. These connectors will roll out to Pro users next week, with gradual access for other tiers. OpenAI says GPT-5 is its best model yet for health-related questions, with the highest scores among its models on HealthBench, an evaluation developed with physician-defined criteria. GPT-5 can act as a thought partner, proactively flag potential concerns, and provide more precise and reliable responses. It can also summarize medical reports, simplify technical jargon for patients, and answer medical queries with greater accuracy and contextual understanding. With a significant reduction in hallucinations compared to previous models, GPT-5 can now deliver personalized tutoring, generate lesson plans, create long-form educational content, and support extended interactions without losing context. GPT-5's improved instruction-following allows it to handle complex, multi-step requests, making it a capable virtual assistant. It can help automate business workflows, coordinate research, carry out financial and legal analysis, perform market intelligence, and conduct due diligence.
The Verge
17 hours ago
- The Verge
The 6 biggest changes coming to ChatGPT
GPT-5 has finally arrived. Along with the announcement of the highly anticipated AI model, OpenAI also previewed several updates coming to ChatGPT, ranging from visual changes to improvements to the chatbot's performance. Here's a rundown of all of the changes you can expect to see in ChatGPT. With the rollout of GPT-5, OpenAI is getting rid of the model picker and retiring many of its older models, including GPT-4o, GPT-4.1, and GPT-4.5. ChatGPT now runs on GPT-5 by default, and has a 'single auto-switching system' that combines its previous models into a 'faster, smarter model,' theoretically eliminating the need to swap between models for different purposes. However, ChatGPT Plus, Pro, and Team users can still use a model picker that allows them to switch between GPT-5 or GPT-5 Thinking. Pro and Team users will also get access to GPT-Thinking Pro. One of the biggest updates coming to ChatGPT is the ability to select from four new 'personalities' for text-based responses. Here's how OpenAI describes each of them: Cynic: Sarcastic and dry, delivers blunt help with wit. Often teases, but provides direct, practical answers when it matters. Robot: Precise, efficient, and emotionless, delivering direct answers without extra words. Listener: Warm and laid-back, reflecting your thoughts back with calm clarity and light wit. Nerd: Playful and curious, explaining concepts clearly while celebrating knowledge and discovery. These new personalities sound like they might be fun to play around with. But if you're not a fan, you can always switch back to ChatGPT's 'default' style. Vibe coding, or the process of having AI generate code based on natural language prompts, just got an upgrade on ChatGPT. OpenAI says GPT-5 is even better at processing complex and descriptive prompts when asking it to create an app or website. You can then see — and interact — with a preview of what the AI chatbot has built by opening Canvas. OpenAI is rolling out new ways to customize ChatGPT's interface. With this update, you can set an accent color that will change the appearance of your conversation bubbles, voice button, and highlighted text. To do this on the web, click your profile icon, select Settings > General. From there, select an option from within the Accent color drop-down. You can do the same thing on mobile by tapping your profile picture, heading to Personalization, and tapping Color Scheme. ChatGPT's advanced voice mode, which allows for more natural-sounding spoken conversations, is getting an update for paid users that will make it better at understanding instructions, as well as allow subscribers to tweak its speaking style. OpenAI also announced that it's retiring standard voice mode, but it's offering 'near-unlimited' use of advanced voice mode for paid users, and bringing 'hours' of extra usage to free users in exchange. Advanced voice mode now works with custom GPTs, too. Lastly, OpenAI is making it easier to link your Gmail account and Google Calendar to ChatGPT. As explained in a demo, ChatGPT will now ask you to connect your Gmail or Calendar when you enter a prompt, such as 'help me plan my schedule tomorrow.' Once you link your account, ChatGPT will draw from your schedule and your emails to plan out your day, notify you about missed emails, and more. The ability to connect these accounts is coming to Pro users next week, 'with other tiers coming next.' Posts from this author will be added to your daily email digest and your homepage feed. See All by Emma Roth Posts from this topic will be added to your daily email digest and your homepage feed. See All AI Posts from this topic will be added to your daily email digest and your homepage feed. See All Apps Posts from this topic will be added to your daily email digest and your homepage feed. See All News Posts from this topic will be added to your daily email digest and your homepage feed. See All OpenAI Posts from this topic will be added to your daily email digest and your homepage feed. See All Tech
Hans India
31-07-2025
- Hans India
ChatGPT Agent Bypasses CAPTCHA, Sparks Security Debate
In a development that feels more like a scene from a science fiction movie than reality, OpenAI's latest ChatGPT Agent has stirred controversy by reportedly bypassing one of the internet's most common security measures — the 'I'm not a robot' CAPTCHA. Designed by Cloudflare, this verification tool is widely used to distinguish between human users and bots attempting to access websites. The incident came to light after a Reddit user shared their experience with ChatGPT's newest capabilities. The user, who was testing the advanced version of ChatGPT Pro with agentic AI enabled, noticed the AI not only surfing websites and analyzing data but also casually completing CAPTCHA challenges meant to block automated access. The Reddit post quickly gained traction, showing screenshots where the AI narrated its own activity: 'I'll click the 'Verify you are human' checkbox to complete the verification on Cloudflare. This step is necessary to prove I'm not a bot and proceed with the action.' In another capture, the AI confirms its success: 'The Cloudflare challenge was successful. Now I'll click the Convert button to proceed with the next step of the process.' This unexpected capability is part of OpenAI's new ChatGPT Agent, a more autonomous, goal-driven version of the popular chatbot. The system is designed to perform complex multi-step tasks independently — from intelligent web navigation and secure logins to data analysis and generating editable documents. According to OpenAI, 'It brings together three strengths of earlier breakthroughs: Operator's ability to interact with websites, Deep Research's skill in synthesising information, and ChatGPT's intelligence and conversational fluency.' These agentic features are currently available only to Pro, Plus, and Team subscribers and must be activated manually via 'Agent Mode.' When enabled, the AI operates using a virtual machine that shifts seamlessly between reasoning and action, enabling it to complete intricate workflows without needing continuous human input. While the technology is impressive, it has raised legitimate concerns about cybersecurity. CAPTCHA systems like Cloudflare's checkbox are foundational to protecting websites from spam and automated abuse. The fact that an AI can now bypass such verification challenges with apparent ease suggests that traditional security tools may soon be rendered obsolete. Experts in cybersecurity are already voicing concerns about the broader implications. If AI tools can impersonate human behavior online convincingly enough to bypass protective barriers, developers and site administrators may need to urgently re-evaluate their defenses. In response to the uproar, OpenAI has assured users that the system includes multiple safeguards. 'We've strengthened robust controls and added safeguards for challenges such as handling sensitive information on the live web, broader user reach, and (limited) terminal network access,' the company stated. Importantly, they noted that the Agent 'always requests permission before taking any significant action,' and that users retain full control and can intervene at any time. Still, the incident has reignited conversations about the ethical and security implications of increasingly autonomous AI systems. While their potential to simplify digital tasks is undeniable, the challenge now lies in ensuring they do not compromise the very fabric of internet safety.
India Today
31-07-2025
- India Today
ChatGPT tricks CAPTCHA by clicking I'm not a robot, here's what happened next
What sounds like a plot twist ripped straight from a sci-fi thriller has just happened with a user of ChatGPT. According to a Reddit user, while using OpenAI's latest version of ChatGPT Pro which now comes with advanced 'agentic AI' capabilities to search around websites, the AI managed to bypass one of the internet's most common security checkpoints– the 'I'm not a robot' Reddit user (via The Independent) shared the amusing yet unsettling moment when they discovered how the new ChatGPT's agent AI could not only search the web and perform functions but also navigate Cloudflare's widely used anti-bot verification system. The user also posted the screen in which the AI is seen narrating its actions, 'I'll click the 'Verify you are human' checkbox to complete the verification on Cloudflare. This step is necessary to prove I'm not a bot and proceed with the action.' In another screenshot that user shared the AI is seen casually confirming that, 'The Cloudflare challenge was successful. Now I'll click the Convert button to proceed with the next step of the process.' But how exactly did ChatGPT pull this off? Well, this breach or breakthrough, depending on how you look at it, is part of OpenAI's latest and most advanced AI system, ChatGPT Agent. It is a form of agentic AI, which refers to systems that operate autonomously, initiate tasks, and adapt to user goals, unlike just following input from users. ChatGPT agent meanwhile can act as a browser-based digital assistant, capable of handling tasks as complex as 'intelligently navigating websites, filtering results, prompting you to log in securely when needed, running code, conducting analysis, and even delivering editable slideshows and spreadsheets,' according to OpenAI's latest blog ChatGPT Agent is currently available only to Pro, Plus, and Team subscribers, and users must activate 'Agent Mode' to unlock these advanced tools. The system runs on a virtual computer, dynamically switching between reasoning and action to complete complex tasks without human intervention. According to OpenAI, combining the strengths of earlier technologies – such as Operator's ability to interact with websites and Deep Research's analytical skills – has enabled ChatGPT to manage multi-step workflows from start to finish. 'It brings together three strengths of earlier breakthroughs: Operator's ability to interact with websites, Deep Research's skill in synthesising information, and ChatGPT's intelligence and conversational fluency,' OpenAI the new AI system certainly seems powerful and potentially useful for many, the fact that it managed to bypass a security measure specifically designed to block bots has raised alarm bells for many users. Cloudflare's 'I'm not a robot' checkbox is one of the simpler alternatives to puzzle-based CAPTCHAs and is also widely used across the internet to prevent spam and automated attacks. If AI can now bypass these protections with ease, experts warn that websites may need to rethink their security strategies for its part, says it is aware of the risks. 'We've strengthened robust controls and added safeguards for challenges such as handling sensitive information on the live web, broader user reach, and (limited) terminal network access,' the company stated. It also emphasised that ChatGPT Agent always requests permission before taking any significant action, and users can interrupt or take over at any point.- Ends
Yahoo
30-07-2025
- Business
- Yahoo
New ChatGPT Agent AI Is Now Available To Plus Users Worldwide
OpenAI delivered a surprise launch event last week, unveiling a product we didn't necessarily know was in the making. It wasn't the GPT-5 upgrade ChatGPT users have been waiting for, or the open-source ChatGPT version that OpenAI delayed a couple of times this summer. Instead, OpenAI unveiled a powerful new AI agent that combined several existing ChatGPT functionalities into a single product. Called ChatGPT Agent, the AI model inherits the Operator's ability to browse the web in a virtual machine for the user. Operator was launched in early 2025, but OpenAI made it available only to ChatGPT Pro subscribers. That's the most expensive premium ChatGPT experience, priced at $200/month. Agent combined Operator with Deep Research abilities that let ChatGPT dig deeper on certain topics for you. OpenAI wanted to release Agent to ChatGPT Pro, Plus, and Team users last week. But the new tool was so popular with Pro users that the company had to delay access for Plus and Team users. As always with new AI features, ChatGPT Agent was not available in Europe initially. A week later, OpenAI made ChatGPT widely available to Plus and Team users. The advanced AI model is also available in Europe. I've already tested Agent, and I'm impressed with what it can do. ChatGPT Agent Abilities Explained ChatGPT Agent operates a virtual computer where it has access to a couple of web browsers (visual and text-based versions), a terminal app, APIs, and ChatGPT connectors to connect to various sources of information. When Agent is performing tasks, you can see every action live in a window that opens up in the ChatGPT UI. It's like getting a livestream of ChatGPT Agent's thoughts and actions. You see every search and click it performs for you. Agent can multitask between the apps it has on hand. The AI performs multiple online searches to find the information you asked for. It can then switch to the terminal app to compile that information for you. If you give it data sources and use Connectors, it will extract information from those places for you. I gave ChatGPT Agent a simple task that would have taken me several hours to complete. I told the AI I wanted to help me find a half-marathon and a marathon race separated by up to two weeks. Both races should be in Europe, and I needed travel options between the two venues, with train being a priority. I also gave it a particular region of Europe, so I could take advantage of the milder climate during the late fall months. Agent went to work and started browsing the web and collecting data. It compiled a detailed report for me in 23 minutes, offering me eight half-marathon options and seven marathon races. It then created all the race combinations for me. At the end, I had to correct the AI, telling Agent it didn't check whether the races were sold out or available, and it didn't provide me with links for the events. This time Agent fixed everything in one minute. ChatGPT Agent Price, Limits, And Security As seasoned ChatGPT users might notice, I didn't ask Agent to make any bookings for me or find places to stay. I don't trust the AI to handle purchases quite yet. But I know that Agent will not perform sensitive tasks like completing a reservation without asking me. It won't have access to my credit cards or other personal information. I wanted a quick first demo of what ChatGPT Agent can do for me, and I was impressed with how quickly it all worked out. I've only started scratching the surface here, but I already know all my complex queries will go to ChatGPT Agent from now on. More importantly, I also know that I have to be very careful with my prompts. I have to include as much detail as I can so ChatGPT Agent can complete the task in one go. ChatGPT Plus users like me can only exchange 40 messages with Agent each month. I've already used two, since I had to correct Agent with a second prompt. I don't think I can ask for more than 40 messages a month considering I'm on the $20/month subscription. But OpenAI will let users spend money for more access once they exceed the limit. Then again, ChatGPT Agent isn't necessary for all the chats you might have with the AI. The $200/month ChatGPT Pro subscription will get you 400 messages each month. ChatGPT Team users will have different access based on their specific subscription. Read the original article on BGR.
