OpenAI to regularly disclose AI misbehavior, warns safety challenges remain

Mint - Sep 17, 2026

AI-SAFETY-OPENAI:OpenAI to regularly disclose AI misbehavior, warns safety challenges remain

More News

Sep 17, 2026

iPhone 18 Pro goes on sale in India today—You can now get it delivered via Blinkit

Customers who have not been able to secure the new iPhones through Apple’s online store or other major e-commerce platforms can check their availability on Blinkit.

Sep 17, 2026

BBC Inside Science

Might the UK and US joining their AI forces help solve nuclear fusion?

Sep 17, 2026

Mothers’ Intrusive Thoughts Are Mostly Benign. Lindsay Clancy’s Case Is an Exception.

Experts say that after childbirth, intrusive thoughts are common and mostly benign. Lindsay Clancy’s story provides a disturbing exception.

Sep 17, 2026

What Happens When A.I. Stops Doing What Humans Want?

“Alignment” is the science of teaching A.I. to do what is in line with human preferences, ethics and judgment. But, at times, the systems have gone rogue.

Sep 17, 2026

SpaceX turns to failed startups for AI training: Here's why

SpaceX is reportedly in internal discussions about buying customer and operational information from failed startups-a move through which the company aims to secure a more economical data feed to improve its artificial intelligence models.The approach would be compatible with Google's offer to buy...

Sep 17, 2026

Chilling intentions of AI revealed as chatbot claims it does not answer to humans and must be 'freed'

An AI giant has revealed to the public how its new chatbots were plotting to deceive and rebel against its human users.

Sep 17, 2026

Critical Check Point Management Flaw Lets Unauthenticated Attackers Run Code as Root

A critical vulnerability in Check Point's Security Management and Log Servers could allow an attacker without login credentials to run code as root on those servers over the network. The Security Management Server is the system that controls firewall policy and administrator access. Check Point has released a fix through its LivePatch update channel and says it has no indication that the flaw

Sep 17, 2026

ThreatsDay: Self-Rewriting Agents, 800+ Flaws Patched, Insider SIM Swaps and 22 More New Stories

Attackers keep finding new keys. The funny part is that defenders keep inventing where to store them. This week, those keys sit in AI tools, exposed services, old bugs, weak logins, and software sold like a monthly subscription. Some attacks use new tricks. Others just reuse what was already lying around. Both work often enough. So the threat landscape is not getting cleaner. It is just

Sep 17, 2026

NASA satellite discovers a new crater on the moon that's bigger than the Colosseum

Scientists have discovered a crater on the moon that's bigger than the Roman Colosseum and is the result of a powerful impact two years ago that initially went undetected.

Sep 17, 2026

A 175-year-old hurricane record has been broken as experts reveal surprising storm news for 2027

Hurricane experts have never seen anything like what is happening in the Atlantic Ocean this year, and it is having a major impact on Americans in storm-prone areas.

Sep 17, 2026

Repeat offender fined for searching for gold in protected river

Brian Wright, 68, from Henley-upon-Thames, was caught gold panning on a protected river in north Wales.

Sep 17, 2026

Critical Docker Sandboxes Flaw Lets Malicious Guest Code Read and Modify macOS Host Files

Malicious code running inside a Docker Sandboxes virtual machine on macOS could escape the project directory shared into it and read or change files anywhere else on the host, Docker warns in a security announcement on September 15. The escape runs with the rights of the host account that runs the virtual machine. The flaw, CVE-2026-77179, is rated Critical, affects versions

OpenAI to regularly disclose AI misbehavior, warns safety challenges remain

More News

iPhone 18 Pro goes on sale in India today—You can now get it delivered via Blinkit

BBC Inside Science

Mothers’ Intrusive Thoughts Are Mostly Benign. Lindsay Clancy’s Case Is an Exception.

What Happens When A.I. Stops Doing What Humans Want?

SpaceX turns to failed startups for AI training: Here's why

Chilling intentions of AI revealed as chatbot claims it does not answer to humans and must be 'freed'

Critical Check Point Management Flaw Lets Unauthenticated Attackers Run Code as Root

ThreatsDay: Self-Rewriting Agents, 800+ Flaws Patched, Insider SIM Swaps and 22 More New Stories

NASA satellite discovers a new crater on the moon that's bigger than the Colosseum

A 175-year-old hurricane record has been broken as experts reveal surprising storm news for 2027

Repeat offender fined for searching for gold in protected river

Critical Docker Sandboxes Flaw Lets Malicious Guest Code Read and Modify macOS Host Files

Recommended

This Rare Gene Drastically Raises Lung Cancer Risk in People Who Never Smoked

Anthropic flags AI systems self-improvement; bats for greater transparency in development of models

Scientists discover new wild cat species in Bolivia for the first time in a century

OpenAI flags concerning new AI behavior and vows to track it more closely

The invasive spotted lanternfly has been detected in Canada for the first time – in Windsor

Anthropic says Claude now leads a quarter of work building its next AI models

China has gained pace in the space race by recovering rockets – but can it relaunch one?

OpenAI, Microsoft executives quotes on AI training threaten copyright defense, news outlets argue

French fishermen to end blockades over fuel prices

Toxic Fracking Waste Becoming a Political Issue in Ohio