Google now lets users toggle off the visible watermark that normally appears on images, videos and music generated with Gemini and its Nano Banana and Omni models. The invisible SynthID watermark and embedded metadata remain in place regardless, meaning the content stays technically traceable as AI-generated even when it no longer looks that way.
Security researchers have confirmed active exploitation of a vulnerability that lets remote attackers log into Macs without a password through a flaw in the screen-sharing feature. Here is how the exploit works, which systems are exposed, and the steps Mac users should take while a patch is developed.
OpenAI and Anthropic have both released cheaper models in recent weeks, a shift analysts attribute to intensifying competition from Chinese AI labs offering comparable performance at a fraction of the cost. The pricing pressure marks a notable change for two companies that have spent years competing primarily on capability rather than cost.
Over the past several weeks, hackers have broken into the industrial control systems of multiple US water utilities in an operation researchers attribute to actors linked to the Iranian government. Here is what has been confirmed about the intrusions, the vulnerable equipment involved, and why water infrastructure keeps turning up as a target.
Aurora Innovation and Kodiak AI have received permits from the California Department of Motor Vehicles to test self-driving trucks on the state's highways, a milestone for an industry that has struggled to gain regulatory footing in one of the country's largest freight markets. Here is what the permits do and do not allow.
Apple has proposed a new structure to a US court that would let it collect fees on digital purchases made through links outside its in-app payment system, and Epic Games says the request goes far beyond what the court has allowed. The dispute is the latest chapter in a years-long legal fight over App Store control.
Rapid revenue growth at Claude maker Anthropic is fuelling hopes that its eventual public listing could become the biggest in history, with some investors floating a valuation as high as $2 trillion. The figures underscore how fast enterprise demand for AI systems has scaled in a short period.
A specialised radiation-blocking vest was tested on a lunar mission, targeting protection at the astronaut's body directly rather than shielding the entire spacecraft. The test succeeded, and researchers say it marks a meaningful step toward protecting crews on longer future Moon and Mars missions.
X is expanding the open-source code behind its 'For You' feed and rolling out new transparency tools that let users see when the platform's ranking systems have affected their account or posts. The move directly addresses years of user speculation about so-called 'shadowbanning.'
Apple sends a push notification directly to a user's lock screen when it identifies state-sponsored spyware targeting their device. Receiving one is rare but serious — security researchers say these alerts are almost never false alarms, and explain exactly what to do if one arrives.
A new tool called ShieldFont lets website owners embed text that reads normally to human visitors but turns into nonsense when scraped by AI training crawlers. The approach adds to a growing toolkit publishers are using to resist having their content harvested for AI models without consent.
For decades, physicists have predicted the existence of glueballs, exotic particles made purely of the force-carrying gluons that bind quarks together, but never definitively observed one. New experimental evidence has produced what researchers call the strongest signal yet that a particle dominated by a glueball component genuinely exists in nature.
Apple's standard iPhone 18 will reportedly not launch alongside the Pro models this fall, arriving instead in early 2027, according to supplier comments spotted by MacRumors. The staggered rollout would continue a shift toward splitting Apple's iPhone launches across the year rather than releasing the full lineup at once.
At the Ai4 conference, three of AI's most respected researchers — Geoffrey Hinton, Fei-Fei Li and Andrew Ng — debated whether open-source access to powerful AI models helps or harms safety. Their exchange captured a widening rift in the field over regulation, openness and how the West can keep pace as China's AI capabilities advance.
Cognition, the AI coding startup behind the Devin software agent, is reportedly in talks to raise new funding at a $40 billion valuation, just months after closing a $1 billion round valuing it at $26 billion. The pace reflects intense investor demand for AI coding tools since Cognition's acquisition of Windsurf.
Google is set to unveil a wave of new Pixel devices at its annual Made by Google event on August 12th. Pre-event leaks have offered an unusually detailed preview of what's expected — from a broader Pixel 11 phone lineup to new wearables and possibly a whole new accessory category for the company.
Researchers report that roadside cameras photographing license plates are increasingly paired with equipment that also captures wireless signals from nearby phones. That combination can turn a simple 'this car passed here' record into a much richer profile of who was actually inside.
Google's Chrome browser has begun rolling out device-bound session credentials, described as among the strongest protections yet against a rapidly growing category of account takeover. The feature targets a technique that steals the session tokens keeping a user logged in, rather than stealing their password.
Google announced that its flagship consumer AI app, Gemini, has surpassed 1 billion users. The company also revealed that 63% of Gemini users talk directly to the assistant using its voice feature, and that Gemini now generates more than 150 million images every day.
On the Riemann hypothesis, one of mathematics' most famous unsolved problems for more than 150 years, an unreleased Anthropic AI model reportedly made significant progress. The model didn't prove the hypothesis, but went further than many experts expected an AI system to get.
Liquid AI's new LFM2.5 model, with just 2.6 billion parameters, is competitive with models four times its size. It's the latest example of a fast-growing small language model movement challenging the assumption that bigger is always better in AI.
One of the rockets forming the backbone of China's space program exploded during launch. If the failure is traced to the YF-100 engine, it could raise broader concerns for a large share of the country's launch fleet.
Valve is gradually extending SteamOS, the operating system built for its Steam Deck handheld, to third-party gaming devices beyond its own hardware. The latest update adds support for outside machines including the Intel-based MSI Claw 8 EX AI+.
After more than a decade, Aptoide has become the first rival app store to return to Google Play in the US, following court-ordered changes that open Android to competing storefronts. Here's what the shift could mean for everyday users.
OpenAI is expanding its Daybreak cyber-defense program and rolling out a new model trained specifically for security tasks, as AI-assisted cyberattacks grow more common. The move comes as AI increasingly plays a central role on both the offensive and defensive sides of cybersecurity.
OpenAI's reported premium smart speaker will reportedly include moving mechanical parts designed to give the device a sense of personality. The company has confirmed the design is not modeled on an Apple product, according to a new report.
To power a new West Texas data center, Amazon is investing in a natural-gas plant that could become one of the largest single sources of greenhouse gas emissions in the United States. The facility will initially operate outside the state's power grid.
AI writing detectors, descended from anti-plagiarism tools, are now casting suspicion on students, writers, and professionals alike. Experts warn the technology is far less reliable than many assume, fueling a broader crisis of trust.
Two decades ago, London's King's Cross was known as one of the city's seediest districts. Today it hosts leading AI labs like DeepMind, illustrating how urban regeneration and tech clustering can reinforce one another.
OpenAI has acquired NextSlide, a startup that used AI to help users build presentations, with the team now moving to work on ChatGPT. Terms of the deal were not disclosed.