Google has been constantly adding new features to the AI Mode in Google Search, and now, the company is introducing a new functionality that offers visual results.

In a blog post, the tech giant said that starting today, Google AI Mode users will be able to “ask a question conversationally and get a range of visual results in AI mode.” For example, if you are searching for design inspiration for your bedroom, AI Mode will help you turn your idea into a vision by showing visuals that match the things you are looking for.

For those not in the loop, Google Search’s AI Mode lets you ask “longer, more complex and nuanced questions”, which would previously require users to make multiple search queries. Initially available in English, Google has since then included support for other languages like Hindi, Brazilian Portuguese, Indonesian, Korean and Japanese.

Google says you can also ask AI Mode for more options and search by simply sharing an image or taking a photo and feeding it to AI Mode. Every image that you see will also have a link associated with it, allowing users to quickly check out a particular item or result.

This is really handy for online shoppers as they can simply describe what they are looking for using natural language and get results without having to use a bunch of filters. Powered by Gemini 2.5’s advanced multimodal and language capabilities, AI Mode can also recognise details and objects in a photo and search for them as well. In case you are wondering, you can also search for a specific thing in an image and ask follow-up questions about it.

However, the feature is currently limited to those living in the United States, but Google is expected to bring it to the rest of the world in the future.

The iPhone 17 series, including the new iPhone Air, has performed well in durability and scratch tests. However, just weeks after launch, several customers and tech experts are reporting that their devices are picking up marks and scratches more easily than expected.

The issue seems to be affecting the Deep Blue, Cosmic Orange and Space Black colourways of the iPhone 17 Pro, iPhone 17 Pro Max and the iPhone Air, with the scratches particularly visible around the sharp edges of the camera bump. In a video, popular YouTuber JerryRigEverything said that the scratches may be because the anodised aluminium layer on the iPhone 17 and the iPhone Pro Max “does not stick to corners very well.”

“For some reason, Apple didn’t add a chamfer, fillet or radius around the camera plateau, and I think it was intentional, so it looks cooler. But that decision to look cool out of the box is going to plague everyone who owns this phone down the road,” he added. As it turns out, if you happen to keep your iPhone 17 Pro or the iPhone 17 Pro Max in the same pocket with everyday items like coins or house or car keys, the metal can chip away at the coating around the camera island, but these items won’t most likely affect the flat rear panel.

Last week, when the iPhone 17 series went on sale, Bloomberg had reported that some of the display models in Apple retail stores already had some scratches on them. The publication also stated that the magnetic MagSafe charger may leave a fairly visible circular mark on the back panel of the iPhone 17 Pro models.

Commenting on the iPhone 17 Pro’s scratchgate controversy, an Apple representative told 9to5mac that the scratches were caused by the worn MagSafe stands that were used in some stores. The tech giant went on to say that they aren’t actually scratches, but the material transfer from the stands to the phone and that they can be removed by cleaning. Apple also claimed that the same issue also affected the iPhone 16 models.

As for the marks on or near the camera bump, Apple says that the camera island’s edges have the same anodised aluminium finish as MacBooks and other iPhone models, but users may see normal wear and tear, like small abrasions over time. iFixit, an online resource for repairing consumer electronics, said that the problem lies in the iPhone 17 Pro’s sharp camera bump, where the anodization process is uneven. Right now, the only way to prevent these scratches from appearing is to put on a case.

A few months ago, OpenAI acquired io Products, the AI hardware startup led by former Apple design chief Jony Ive, and was reportedly working on several AI-powered gadgets.

Now, a new report by The Information, citing people “with direct knowledge of the matter”, claims that the ChatGPT developer is considering making a device that “resembles a smart speaker without a display”. It goes on to say that OpenAI has already secured contracts with Luxshare and Goertek – two Apple suppliers who make parts for iPhone, AirPods, HomePods and Apple Watches for its speakers.

As it turns out, OpenAI and io Products may also be working on smartglasses, a digital voice recorder and a wearable Humane-like “pin”. The publication also claims that these devices will launch sometime in late 2026 or early 2027 and will have OpenAI branding.

OpenAI CEO Sam Altman had previously said that one of the devices would be a pocket-sized, contextually aware screenless device, but the report now hints at other form factors like eyewear. However, as previous rumours suggested, it looks like there won’t be any in-ear devices. The Information also says that OpenAI has started using Apple’s supply chain in China to make these devices and that the Cupertino-based tech giant cancelled a meeting with its suppliers in China to prevent OpenAI from poaching talent from Apple.

In the last few months, Apple employees who were part of the company’s AI division have jumped ship to OpenAI ever since the Sam Altman-led company announced its partnership with the original iPhone’s designer ,Jony Ive. Tang Tan, Apple’s former design head, left the tech giant to become OpenAI’s chief hardware officer.

However, OpenAI is yet to show a prototype of any of their devices and hasn’t said anything about form factor and price either, but if true, this might mean that we can see OpenAI taking inspiration from Apple for its debut products.

The Trump administration said on Friday it would ask companies to pay $100,000 per year for H-1B worker visas, prompting some big tech companies to warn visa holders to stay in the U.S. or quickly return.

The change could deal a big blow to the technology sector that relies heavily on skilled workers from India and China.

Since taking office in January, Trump has kicked off a wide-ranging immigration crackdown, including moves to limit some forms of legal immigration. The step to reshape the H-1B visa program represents his administration’s most high-profile effort yet to rework temporary employment visas.

“If you’re going to train somebody, you’re going to train one of the recent graduates from one of the great universities across our land,” said Commerce Secretary Howard Lutnick. Train Americans. Stop bringing in people to take our jobs.”

Trump’s threat to crack down on H-1B visas has become a major flashpoint with the tech industry, which contributed millions of dollars to his presidential campaign.

Microsoft, JPMorgan and Amazon responded to the announcement by advising employees holding H-1B visas to remain in the United States, according to internal emails reviewed by Reuters.

They advised employees on the H-1B visas who were outside the U.S. to return before midnight on Saturday (0400 GMT on Sunday), when the new fee structures are set to take effect.

“H-1B visa holders who are currently in the U.S. should remain in the U.S. and avoid international travel until the government issues clear travel guidance,” read an email sent to JPMorgan employees by Ogletree Deakins, a company that handles visa applications for the U.S. investment bank.

Microsoft, JPMorgan, law firm Ogletree Deakins, which represents the bank on the issue, and Amazon did not immediately respond to Reuters requests for comment.

Critics of the H-1B program, including many U.S. technology workers, argue that it allows firms to suppress wages and sideline Americans who could do the jobs. Supporters, including Tesla CEO and former Trump ally Elon Musk, say it brings in highly skilled workers essential to filling talent gaps and keeping firms competitive. Musk, himself a naturalized U.S. citizen born in South Africa, has held an H-1B visa.

Some employers have exploited the program to hold down wages, disadvantaging U.S. workers, according to the executive order Trump signed on Friday.

The number of foreign science, technology, engineering and mathematics (STEM) workers in the U.S. more than doubled between 2000 and 2019 to nearly 2.5 million, even as overall STEM employment only increased 44.5% during that time, it said.

MOVE COULD DETER GLOBAL TALENT

Adding new fees “creates disincentive to attract the world’s smartest talent to the U.S.,” said Deedy Das, partner at venture capital firm Menlo Ventures, on X. “If the U.S. ceases to attract the best talent, it drastically reduces its ability to innovate and grow the economy.”

The move could add millions of dollars in costs for companies, which could hit smaller tech firms and start-ups particularly hard.

Reuters was not immediately able to establish how the fee would be administered. Lutnick said the visa would cost $100,000 a year for each of the three years of its duration but that the details were “still being considered.”

Under the current system, entering the lottery for the visa requires a small fee and, if approved, subsequent fees could amount to several thousand dollars.

Some analysts suggested the fee may force companies to move some high-value work overseas, hampering America’s position in the high-stakes artificial intelligence race with China.

“In the short term, Washington may collect a windfall; in the long term, the U.S. risks taxing away its innovation edge, trading dynamism for short-sighted protectionism,” said eMarketer analyst Jeremy Goldman.

INDIA ACCOUNTS FOR MOST H-1B VISAS

India was the largest beneficiary of H-1B visas last year, accounting for 71% of approved beneficiaries, while China was a distant second at 11.7%, according to government data.

In the first half of 2025, Amazon.com and its cloud-computing unit, AWS, had received approval for more than 12,000 H-1B visas, while Microsoft and Meta Platforms had over 5,000 H-1B visa approvals each.

Lutnick said on Friday that “all the big companies are on board” with $100,000 a year for H-1B visas.

“We’ve spoken to them,” he said.

Many large U.S. tech, banking and consulting companies declined to comment or did not immediately respond to requests for comment. The Indian embassy in Washington and the Chinese Consulate General in New York also did not immediately respond to requests for comment.

Shares of Cognizant Technology Solutions, an IT services company that relies extensively on H-1B visa holders, closed down nearly 5%. U.S.-listed shares of Indian tech firms Infosys and Wipro closed between 2% and 5% lower.

IMMIGRATION CRACKDOWN

Aaron Reichlin-Melnick, policy director of the American Immigration Council, questioned the legality of the new fees. “Congress has only authorized the government to set fees to recover the cost of adjudicating an application,” he said on Bluesky.

The H-1B program offers 65,000 visas annually to employers bringing in temporary foreign workers in specialized fields, with another 20,000 visas for workers with advanced degrees.

Nearly all the visa fees have to be paid by the employers. The H-1B visas are approved for a period of three to six years.

Trump also signed an executive order on Friday to create a “gold card” for individuals who can afford to pay $1 million for U.S. permanent residency.

Microsoft Inc said on Tuesday that it seized nearly 340 websites tied to a rapidly growing Nigerian-based service that allowed users to carry out phishing operations that stole at least 5,000 Microsoft user credentials. Microsoft obtained an order from the U.S. District Court in Manhattan earlier this month to seize domains associated with Raccoon0365, the subscription service that allowed users to carry out massive phishing campaigns, which sometimes involved thousands of emails at a time, according to Steven Masada, assistant general counsel for Microsoft’s Digital Crimes Unit.

Raccoon0365’s service, which operates through a private Telegram channel with more than 850 subscribers, enables users to impersonate trusted brands and get targets to enter Microsoft login credentials on phony Microsoft login pages, Masada said in a blog posted on Microsoft’s website. The service has generated for its small group of operators at least $100,000 in cryptocurrency payments since launching in July 2024, Masada said in the blog.

Microsoft said the seizure of the websites occurred over a period of days earlier this month. Microsoft identified Nigeria-based Joshua Ogundipe as the leader and main operator of Raccoon0365. Ogundipe did not immediately respond to an email request for comment sent to the email address identified by Microsoft in its court filing. “Cybercriminals don’t need to be sophisticated to cause widespread harm,” Masada said. “Simple tools like Raccoon0365 make cybercrime accessible to virtually anyone, putting millions of users at risk.” Raccoon0365 subscribers have targeted a wide swath of industries, Masada said, and separate court filings allege that “a significant portion” of Raccoon0365 activity targets organizations based in New York City.

Masada said Microsoft identified what it said was a Raccoon0365-related effort using tax-themed phishing emails to target more than 2,300 organizations, mostly in the U.S., between February 12 and February 28 this year, according to a company blog posted in April. Errol Weiss, chief security officer of the Health Information Sharing & Analysis Center (Health-ISAC), which provides cybersecurity services to member health organizations and is a co-plaintiff alongside Microsoft, said Raccoon0365 has been linked to successful credential harvesting through phishing campaigns at at least five unnamed healthcare organizations, while targeting 25 health sector organizations overall.

Once hackers gain that access, any number of things can happen, Weiss said. “So many of the attacks start because somebody gave up their user name and password to a bad guy,” Weiss said in an interview. “Once that cybercriminal has access to the network, then it’s just up to the imagination in terms of what comes next and how they monetize it.” The Raccoon0365 operators used services provided by Cloudflare to help hide the service’s backend infrastructure, the internet services firm said in its own blog post. Cloudflare worked with Microsoft and the U.S. Secret Service to disrupt Raccoon0365 operations on its platform and prevent the operators from establishing new accounts, the company said. Blake Darché, the head of threat intelligence at Cloudflare, said in an interview that the Raccoon0365 operators made some key operational security mistakes but were highly effective. “They’re in people’s accounts, they compromise lots of people, and it needs to obviously be stopped,” he said.