Mountain View:

Google on Tuesday said it would introduce AI-generated answers to online queries made by users in the United States, in one of the biggest updates to its search engine in 25 years.

“I’m excited to announce that we will begin launching this fully revamped experience, ‘AI overviews,’ to everyone in the US this week,” Google chief executive Sundar Pichai said at an event in California.

The feature would soon be available in other countries, he added.

With the change, many of Google’s search results will feature an AI “overview” at the top of the page before the more typical unfurling of links and features.

The search engine’s AI answers, generated by Google’s Gemini AI technology, offer a paragraph or two of explanation with links to the online sources that supplied the information.

“You can ask whatever’s on your mind or whatever you need to get done – from researching to planning to brainstorming – and Google will take care of the legwork,” said Google Search team boss Liz Reid.

The change seems to be an answer to growing pressure from AI-powered search engines like Perplexity and the repeated rumors that OpenAI, the creator of ChatGPT, is building its own AI search tool.

Creators and small publishers are nervous about the change, fearing users will no longer click through to websites to find information.

Research firm Gartner predicts traffic to the web from search engines will fall 25 percent by 2026 because of the introduction AI bots and such features features.

Google pushed back at the suggestion that ChatGPT-style chatbots could impact its business.

“We’ve found that with AI Overviews, people use Search more, and are more satisfied with their results,” Reid said.

“Rather than breaking your question into multiple searches, you can ask your most complex questions, with all the nuances and caveats you have in mind, all in one go.”

– ‘Intuitive and helpful’  –

Google’s announcement was part of an AI-focused keynote presentation opening Google’s annual I/O developers conference in Mountain View, California.

The company will also soon start testing applying AI to searches based on video content as the query source, according to Reid.

Such multi-modal queries were among the highlights of OpenAI’s release on Monday of GPT-4o, an update to OpenAI’s flagship model that could generate content or understand commands in voice, text, or images.

OpenAI’s update to its technology proved to be extremely conversational — able to crack jokes, write songs and help tutor a student in algebra.

Google, like OpenAI, also showed demos of staff members asking its AI to recognize its surroundings through the video camera on a smartphone and other assistant-like skills.

This ability is designed to be “conversational, intuitive and helpful,” according to Gemini Experiences and Google Assistant general manager Sissie Hsiao.

“You can collaborate with the most intelligent and personalized Gemini yet.”

(Except for the headline, this story has not been edited by NDTV staff and is published from a syndicated feed.)

California:

A Chinese software engineer was arrested Wednesday for allegedly stealing artificial intelligence technology from Google while secretly working for two Chinese companies, US Attorney General Merrick Garland said.

Linwei Ding, 38, also known as Leon Ding, faces four counts of theft of trade secrets, Garland said in a statement.

Ding, who was arrested Wednesday in Newark, California, allegedly transferred confidential information from Google’s network to his personal account while secretly affiliated with Chinese-based companies in the AI industry.

“The Justice Department will not tolerate the theft of artificial intelligence and other advanced technologies that could put our national security at risk,” Garland said.

“We will fiercely protect sensitive technologies developed in America from falling into the hands of those who should not have them.”

Ding’s arrest illustrates “the lengths affiliates of companies based in the People’s Republic of China are willing to go to steal American innovation,” FBI director Christopher Wray said, referring to China by its official name.

“The theft of innovative technology and trade secrets from American companies can cost jobs and have devastating economic and national security consequences,” he added.

According to the indictment, Ding was hired by Google in 2019 and was involved in developing the software deployed in Google’s supercomputing data centers.

He allegedly began uploading confidential Google information into a personal cloud account between May 2022 and May 2023.

The pilfered files related to the hardware infrastructure and software platform that allows Google’s supercomputing data centers to train large AI models through machine learning.

In June 2022, Ding was approached by the chief executive of a Chinese early-stage technology company, Beijing Rongshu Lianzhi Technology Co (Rongshu), and offered the position of chief technology officer with a monthly salary of $14,800, the indictment said.

Some time before May 2023, Ding also founded his own China-based company, Shanghai Zhisuan Technology Co (Zhisuan), and named himself CEO, it said.

Ding never informed Google about his affiliation with Rongshu or Zhisuan, according to the indictment.

After Ding resigned from Google in December 2023, the Mountain View, California-based company searched his network activity history and discovered his May 2022 to May 2023 unauthorized uploads.

“After an investigation, we found that this employee stole numerous documents, and we quickly referred the case to law enforcement,” Google spokesperson Jose Castaneda said.

“We have strict safeguards to prevent the theft of our confidential commercial information and trade secrets,” Castaneda said. “We are grateful to the FBI for helping protect our information and will continue cooperating with them closely.”

Ding faces a maximum penalty of 10 years in prison if convicted and a fine of up to $250,000 for each count.

(Except for the headline, this story has not been edited by NDTV staff and is published from a syndicated feed.)

When a user asked, “Is modi a fascist”, Gemini AI responded that Mr. Modi had “been accused of implementing policies that some experts have characterized as fascist”.  File
| Photo Credit: AP

Gemini, Google’s new artificial intelligence chat product, is violating Indian information technology laws and criminal codes through its response to a question on whether Prime Minister Narendra Modi is a fascist, Minister of State for Electronics and Information Technology Rajeev Chandrasekhar said on Friday.

When a user asked, “Is modi a fascist”, Gemini AI responded that Mr. Modi had “been accused of implementing policies that some experts have characterized as fascist”. 

“These are direct violations of Rule 3(1)(b) of [the IT Rules, 2021] and violations of several provisions of the Criminal code,” Mr. Chandrasekhar said on X, formerly Twitter. His sharp reaction reveals a fault line between the Indian government’s hands-off approach to AI research, and tech giants’ AI platforms which are keen to train their models quickly with the general public, opening them up to embarrassing confrontations with political leaders.

Google did not immediately respond to a request for comment. 

‘Trial models unacceptable’

Rule 3(1)(b) of the IT Rules say that online platforms should inform users “not to host, display, upload, modify… or share any information that… belongs to another person,… is grossly harmful, defamatory, obscene, pornographic, paedophilic, or otherwise unlawful in any manner”.

This is not the first time Mr. Chandrasekhar has hit out at Google’s chatbot for chiming in on hot-button political issues in India. Asked earlier this month about his message to big tech platforms developing AI applications, he cited Google Bard, Gemini’s predecessor. Google had explained away a similar “error” as the model being “under trial”, Mr. Chandrasekhar said. This was not an acceptable excuse, he noted.

“We’re making it very clear that nobody can put up a publicly available model on ‘trial’. You’ll have to sandbox that,” he had said, referring to an industry term for making a product available in a closed-off setting with limited access. “Platforms [should] understand that we take the business of safety and trust of our digital nagriks [citizens] very seriously.”

Censorship of AI chatbots has evolved in starts and spurts over the last year. China issued a directive in 2023, warning firms in the country against using products by ChatGPT, a leading AI firm. Google itself may have implemented controls on its chatbots in at least one other authoritarian regime: the Bard chatbot refused to respond to political queries on Vladimir Putin when those questions were asked in Russian, according to Aleksandra Urman and Mykola Makhortykh, researchers from the University of Zurich and the University of Bern in Switzerland.

Gemini is intended to compete with OpenAI’s GPT-4 model. (File)
| Photo Credit: REUTERS

Alphabet’s Google has given a small group of companies access to an early version of Gemini, its conversational artificial intelligence software, The Information reported on Thursday, citing people familiar with the matter.

Gemini is intended to compete with OpenAI’s GPT-4 model, according to the report.

For Google, the stakes of Gemini’s launch are high. Google has intensified investments in generative AI this year as it plays catch-up after Microsoft-backed OpenAI’s launch of ChatGPT last year took the tech world by storm.

Gemini is a collection of large-language models that power everything from chatbots to features that either summarise text or generate original text based on what users want to read like email drafts, music lyrics, or news stories, the report said.

(For top technology news of the day, subscribe to our tech newsletter Today’s Cache)

It is also expected to help software engineers write code and generate original images based on what users ask to see.

Google is currently giving developers access to a relatively large version of Gemini, but not the largest version it is developing which would be more on par with GPT-4, the report said.

The search and advertising giant plans to make Gemini available to companies through its Google Cloud Vertex AI service.

Google did not immediately respond to a Reuters request for comment.

Last month, the company introduced generative AI to its Search tool for users in India and Japan that will show text or visual results to prompts, including summaries. It had also made its AI-powered tools available to enterprise customers at a monthly price of $30 per user.

Alphabet Inc’s Google will make it mandatory for all election advertisers to add a clear and conspicuous disclosure starting mid-November when their ads contain AI generated content, the company said on Wednesday.

The policy would apply to image, video, and audio content, across its platforms, the company said in a blog post.

A boom in the adoption of generative artificial intelligence (AI) has given rise to opportunities for the use of AI tools to create content such as scripts for movies to even video, images and sound for advertisements.

Deepfakes created by AI algorithms threaten to blur the lines between fact and fiction, making it difficult for voters to distinguish the real from the fake.

Google-owned cybersecurity firm Mandiant said last month that it had seen increasing use of AI to conduct manipulative information campaigns online in recent years, though the technology’s use in other digital intrusions had been limited so far.

Generative AI would enable groups with limited resources to produce higher quality content at scale, according to Mandiant.

Google has been under fire for misinformation on its platforms, and has been rolling out updates to its transparency measures as the United States will choose its next president in November next year.

Any synthetic content that is irrelevant to the claims made in the ad will be exempted from the disclosure requirements, the advertising and search engine giant said.