What´s new in ChatGPT-5

OpenAI´s GPT-5 arrives as a free consumer model that promises faster, smarter interactions and new tools for creative, coding and productivity tasks in generative Artificial Intelligence.

OpenAI has launched GPT-5 and made it available to consumers as the backbone of the newest ChatGPT release, the article reports. The model is presented by the company as ´our smartest, fastest, most useful model yet, with built-in thinking that puts expert-level intelligence in everyone’s hands´, a positioning that underlines both ambition and accessibility. The piece, by Nikolas Kokovlis and published on 8 august 2025, frames GPT-5 as an evolutionary step that bundles several interface and capability changes intended to simplify how people interact with large language models.

Among the headline changes is a ´real time router´ that automates model selection so users no longer have to pick between specialist reasoning models and faster general models. ChatGPT will also offer four preset personalities — ´cynic´, ´robot´, ´listener´ and ´nerd´ — to steer tone and behaviour. OpenAI says GPT-5 reduces hallucinations, reporting it was 45% less likely to contain a factual error than GPT-4o in web-search-enabled tests with anonymised prompts, and 80% less likely to contain an error than the older o3 model when ´thinking´; despite those gains the company and observers acknowledge hallucinations are not eradicated, and the model may still err roughly one in ten times on common tasks. Writing and coding are highlighted improvements: OpenAI claims more resonant prose and an enhanced ´vibe coding´ workflow for creating websites, apps and games from single prompts. GPT-5 also adds agentic productivity links, integrating with Gmail and Google Calendar for scheduling and follow-ups, though that feature will be initially limited to paid ´pro´ users. Healthbench performance is higher, and OpenAI frames the model as a more proactive partner for health queries while stressing it does not replace medical professionals.

The article cautions that real-world judgement will follow broader testing. Early external testers quoted by reuters said the jump from GPT-4 to GPT-5 feels smaller than the leap from GPT-3 to GPT-4, and sceptical experts such as gary marcus suggested fans will find reasons to celebrate while noting the release is not the transformative breakthrough some expected. In short, GPT-5 packages several practical improvements and new user-facing features; whether it represents an incremental refinement or a substantive advance will become clearer as independent evaluations and everyday use accumulate.

76

Impact Score

Fraud controls don’t guarantee consumer trust

Over a third of companies deploy Artificial Intelligence to fight fraud, but Experian finds rising losses, shifting threats and a credibility gap between business controls and consumer trust.

###CFLOAD###

Contact Us

Got questions? Use the form to contact us.

Contact Form

Clicking next sends a verification code to your email. After verifying, you can enter your message.

Please check your email for a Verification Code sent to . Didn't get a code? Click here to resend