Technologies

Hackers Are Trying to Copy Gemini via Thousands of AI Prompts, Google Reports

A new report from Google warns the industry of attempts to clone AI models.

In a new Threat Tracker report published Thursday, Google said hackers are engaging in «distillation attacks,» including one case in which they used more than 100,000 AI prompts to steal the company’s technology for its Gemini AI model.

Google said the attacks seem to be coming from adversaries in countries including North Korea, Russia and China, and that the attempts to steal AI intellectual property and likely clone it into AI models in other languages are part of a broader set of AI-based attacks and malware the company has seen emerge.

The company identifies these attempts as model extraction attacks, which, it says, «occur when an adversary uses legitimate access to systematically probe a mature machine learning model to extract information used to train a new model.»

That could mean using AI to flood Gemini with thousands of prompts to replicate its model capabilities. Google noted in the report that this is not a threat to its users, but rather to service providers and model builders, who could be vulnerable to having their work stolen and replicated.

AI competition and AI thievery

John Hultquist, chief analyst for the Google Threat Intelligence Group, which put together the report, told NBC News that Google may be one of the first companies to face these types of theft attempts, but there could be many more. «We’re going to be the canary in the coal mine for far more incidents,» he said.

The war over AI models has intensified on several fronts, most recently withs Chinese companies such as ByteDance introducing advanced video generation tools. Last year, Chinese AI company DeepSeek rattled the AI industry, which had been primarily led by US companies, by introducing a model that rivaled the world’s top AI technology. OpenAI later accused DeepSeek of training its AI on existing technology in ways similar to those described by Google in its new report.

(Disclosure: Ziff Davis, CNET’s parent company, in 2025 filed a lawsuit against OpenAI, alleging it infringed Ziff Davis copyrights in training and operating its AI systems.)

Technologies

AT&T Says It’s Pumping $250 Billion Into New Infrastructure Improvements

Technologies

Apple’s New Smart Home Display Delayed Until Fall Over Siri Issues

It has been nearly a year and a half since the company announced the AI-powered product.

Your home could get smarter with Apple’s Siri, but it will have to wait a few more months. Bloomberg reported the iPad-shaped AI home hub won’t be ready until September, several months after the company was hoping to launch it this spring. Apple engineers first need to complete work on a new and improved Siri assistant for the home device, code-named J490, according to Bloomberg.

Apple was hoping to release J490 this month, along with a slew of other new devices, including the iPhone 17e, MacBook Neo, MacBook Air M5, new Pro models, and iPad Air M4. Apple first teased the smart home display in November 2024.

A representative for Apple did not immediately respond to a request for comment.

Siri is Apple’s virtual assistant that uses voice recognition and AI to fulfill a variety of tasks and commands, along with intriguing uses. You might use Siri to find your iPhone — «Hey Siri, where are you?» — or to hear the weather forecast — «Siri, what will the weather be today?» Siri is available on iPhones, MacBooks and iPads. It was launched in 2011 as a feature of the iPhone 4S.

As CNET reported last month, Apple engineers have struggled to push the upgraded Siri assistant out the door. It isn’t fast enough, gets confused by complex commands and doesn’t interact well with other Apple AI models. The company is also wrestling with how much personal data to access to inform the AI, and the new Siri is not yet able to complete in-app tasks, such as finding a photo and posting it to socials, all with one command.

It has been nearly two years since Apple announced that it would give Siri a major upgrade. In the meantime, competitors like Alexa Plus and Gemini for Home have entered the marketplace.

Tech tester Jon Rettinger, whose YouTube channel has 1.66 million subscribers, says the repeated delays in upgrading Siri can «erode» confidence in Apple’s ability to keep up in the AI race.

«Apple as a whole is still one of the strongest companies on the planet. But their AI play is clearly the weakest link in an otherwise very strong chain,» Rettinger told CNET.

Rettinger said he has had issues getting Siri to complete basic commands, such as setting two alarms at the same time, and that it’s a bit of «a mess» right now.

«Having said that, the iPhone has such massive market penetration that I’m not sure it will actually matter in the end. Which is kind of wild when you think about it,» Rettinger said.

Facial recognition for residents

The hardware for the forthcoming smart home display has already been finished. It resembles an iPad and can be either attached to a wall or rest on a half-domed-shaped base, the Bloomberg report said.

The device will be equipped with facial recognition, so when residents walk up to it, they will be shown personalized data such as music preferences, news headlines, appointments, reminders, tasks and so on.

The screen interface will include a bunch of circular app icons, similar to the display on an Apple Watch. The Bloomberg report said the smart home display will be the first of several home devices by Apple. Future products include a tabletop robotic limb with a 9-inch screen, a smart security camera and a Face ID-enabled smart doorbell.