Google deploys Gemini Nano in Chrome to protect users from online scams
Google is deploying advanced AI tools across Search, Chrome, and Android to detect and block online scams more effectively.
The tech giant announced on Wednesday, May 8, that it will be integrating its on-device large language model (LLM) called Gemini Nano with the desktop version of its Chrome browser to act as an additional layer of defence against online scams. It will be expanded to Android devices in the future.
Gemini Nano will be used to improve Enhanced Protection mode of Safe Browsing on Chrome.
'The on-device approach provides instant insight on risky websites and allows us to offer protection, even against scams that haven't been seen before. Gemini Nano's LLM is perfect for this use because of its ability to distill the varied, complex nature of websites, helping us adapt to new scam tactics more quickly,' Google said in a blog post.
The company said it was already using the AI-enhanced approach to address remote tech support scams.
Aside from Gemini Nano, Google said it will start displaying AI-powered warnings to Chrome and Android users. 'When Chrome's on-device machine learning model flags a notification, you'll receive a warning with the option to either unsubscribe or view the content that was blocked. And if you decide the warning was shown incorrectly, you can choose to allow future notifications from that website,' the company said.
Google has been leveraging AI to detect and block online scams for several years now. For instance, AI is used to block millions of scammy search results on Google Search every day, as per the company.
These AI-powered systems are used to scan vast quantities of text on the web, identify coordinated scam campaigns, and detect emerging threats. 'For example, we've observed a significant increase in bad actors on the web impersonating airline customer service providers and scamming people in need of help,' Google said, adding that AI has helped the company reduce these types of airline customer service scams by more than 80 per cent in Search.
AI has also reportedly enabled the big tech company to identify 20 times more scam web pages than before. To address scam calls and messages, Google said it has undertaken efforts such as rolling out on-device, AI-powered Scam Detection in Google Messages and Phone by Google on Android devices.
Try Our AI Features
Explore what Daily8 AI can do for you:
Comments
No comments yet...
Related Articles
Mint
an hour ago
- Mint
The best new features coming to your iPhone, iPad and Mac (and what's missing)
Apple software boss Craig Federighi led the reveal of fall updates coming to iPhones, iPads, Macs and more. A spam call screener, live text translation, an AI workout coach and more. At Monday's Worldwide Developers Conference keynote, Apple announced enhancements throughout its system apps and a new software look called 'Liquid Glass," coming soon to compatible iPhones, iPads, Apple Watches, Macs and other devices. The bad news: iPhone XS and XR owners will need to upgrade this fall to get the latest software. The current iOS 18 will graduate to…iOS 26. The numbers attached to iOS, iPadOS and MacOS will now represent years. Like carmakers, Apple is using the coming year, not the one we're currently in. (I guess that makes sense?) What really matters is, as usual, free software upgrades are coming to most of your Apple devices around the time the new iPhones appear. The biggest change is the redesign to all of Apple's operating systems. Buttons are rounder. Menus are more transparent. The glass effect means text and other interface elements reflect as you tilt your phone. The idea is to make the software look the same across devices. The reality is, many people will have to relearn how to use some parts of their phone. We can help with that when the time comes! Your Apple devices are getting a fresh coat of paint the company calls 'Liquid Glass.' Expect more transparent buttons, menus and windows—and a new home for some settings to learn. And as for the elephant in the room, Apple Intelligence: The company announced AI features throughout the keynote, tucked into various apps and services, as you'll see below. But we didn't hear any details on the delayed Siri overhaul announced last year. Messages gets WhatsApp tricks: Typing indicators in group chats! At last! You can also add backgrounds for your text threads, and create polls—so your group of friends can easily vote on the best destination for an outing. Apple Intelligence can even suggest polls based on the conversation, for instance, if someone asks, 'Where should we stay this weekend?" Though iPhone-Android conversations improved last year, these upgrades apply only to blue-bubble, iPhone-only threads. Live translation: Apple now does on-device live translation, taking a page from Google and Samsung. In Messages, this means texts will be delivered in the recipient's preferred language. In FaceTime, the translations will appear as subtitles above the video call. And on a phone call, you'll hear the translations live as you speak with the other person. It may require the patience of both participants, though. In the demo, there was a delay, as the software worked on captioning the live speech. You will be able to see live translated captions in FaceTime video and phone calls. Unwanted calls and texts: Spam is so prevalent, I now have a burner number to deal with the onslaught. Apple has new tricks to help manage these pesky reach-outs. When you receive a call from an unknown number, iOS 26 can automatically answer the call. Once the caller shares their name and reason for the call, the phone will start to ring and you'll see a summary of the screened call on the lock from most unknown senders will be silenced and tucked away in a separate inbox automatically. But urgent texts, such as verification codes or restaurant wait-list updates, will still go to your main list. Hold Assist: Finally, an Android-favorite feature is coming to save you from waiting for your airline or insurance customer service. If the software detects hold music, you can tap a new hold option and the call will stay connected, but you can do other things. Once the agent is available, your phone will ring and the software will tell the agent you'll be there shortly. Hopefully the agent won't mind waiting on hold a few seconds. On-screen Visual Intelligence: This image-interpreting AI was limited to camera captures on compatible devices. Now, when you take a screenshot, you get a new set of options around the frame. You can look up a lamp that is shown in the image and even get a link to buy it in a shopping site like Etsy or Anthropologie. You can create a calendar event based on information from an online concert poster. You can also ask ChatGPT about a celebrity that appears in a shot on your social-media app. More powerful Mac search: The Mac's best-kept secret is Spotlight—aka that floating search bar that appears when you hit Command+Space. You can now do more things right inside the search environment, such as create a calendar event or start an audio recording. You can also use shortcuts, like 'sm" for 'send message" and 'ar" for 'add reminder." More iPhone on the Mac: As part of Apple's move to keep you buying its products, the Mac can now mirror more iPhone features. Waiting for an Uber Eats delivery? You can now see the same live phone updates on your laptop screen. And the redesigned phone app is also coming to the Mac, so you can manage calls from the larger screen—as long as you enable the right settings. You can open and resize more windows on the iPad, which is getting a slew of multitasking enhancements. iPad gets more Mac-like: Apple has been touting the iPad as a Mac replacement for years, but the tablet never quite measured up. With a menu bar up top and the ability to open and resize more windows, it may now be a lot closer. The Preview app, previously exclusive to the Mac, is also coming to iPad for PDF editing. AI workout buddy: On the Apple Watch, you can launch a generative-AI coach, trained on voice data from Fitness+ trainers and informed with your own workout history. It can pepper your workout with encouragement ('Second run this week. You're crushing it!") and update you on pace, distance and heart rate. Workout Buddy will be an option for popular exercise types, such as runs, rides and strength training. I am looking forward to seeing how this AI coach responds to…my slowest jog ever. One year ago, Apple used this same occasion to announce a slew of AI-powered features. But Siri was hardly mentioned Monday. In March, the company said the voice assistant upgrade is taking longer than expected, and it pulled an ad showing a more personalized Siri capability. Apple will, however, let app makers tap in to its on-device large language models. Executives say the AI technology will be 'fast, private and offline." App developers, such as AllTrails, will now have access to Apple's on-device models to power AI features. One example is a new chat experience in the hiking app AllTrails. You can type what kind of experience you're looking for, and the app will tap in to Apple's AI models to generate a route suggestion, even if you're off the grid. We won't get the full scope of possible applications until developers have time to build them, but it's one way Apple is incorporating more AI into its devices. Meanwhile, Apple's competitors won't shut up about AI search, agentic operators, video generation or deep research. Maybe Apple will be all over that…next year? Write to Nicole Nguyen at
Hindustan Times
an hour ago
- Hindustan Times
AI must be subordinate to fairness, equity, human dignity: Justice Surya Kant
Supreme Court judge Justice Surya Kant has said justice cannot be reduced to a digital product, warning that artificial intelligence (AI) must always remain subordinate to fairness, equity, and human dignity. 'Justice, unlike software, is not a product to be optimised, but a principle to be honoured. Technology must remain subordinate to our higher commitments to fairness, equity, and human dignity,' said Justice Kant, who is set to become the Chief Justice of India in November. Speaking at Microsoft's Fireside Chat on 'AI and Law' on June 6, Justice Kant cautioned that while AI promises to enhance access, efficiency, and transparency in the legal system, unchecked deployment could mirror and even magnify existing societal inequities. 'Technology, if left unchecked, can reflect and reinforce societal inequities. AI is not a perfect technology, and it can perhaps never replace the human element that the entire Rawlsian theory of justice hinges on,' he said. Rawlsian theory refers to the philosophy of justice developed by John Rawls, an American political philosopher. The core of the theory is the concept of 'justice as fairness', which aims to reconcile the seemingly competing values of freedom and equality. Justice Kant acknowledged the global nature of the challenges AI presents, particularly issues like algorithmic bias, hallucinated legal citations, and data protection. 'Take, for instance, the fictitious legal precedents that chatbots routinely come up with when faced with complex legal propositions,' said Justice Kant, warning of the risks of relying blindly on AI in sensitive domains like law. He spoke about growing cyber threats to courts and the judiciary, including ransomware attacks and doxing of judges, and said such digital risks were now 'a matter of constitutional resilience.' He said India has responded proactively, with secure e-filing platforms, the National Judicial Data Grid, and virtual hearings backed by multi-layered authentication. 'Cybersecurity is not a matter of IT hygiene, but of constitutional resilience…courts must invest not just in secure infrastructure, but in public confidence,' Justice Kant said. Justice Kant said the adoption of AI must not be driven by novelty or efficiency alone. 'We do so not as passive observers, but as stewards of a future we must shape with wisdom and purpose… Shaping the future demands more than innovation—it calls for an unwavering adherence to foundational values.' Justice Kant said India's judicial digital transformation, while ambitious, is being shaped through collaboration between technologists, judges, civil society, and academics via a dedicated Centre for Research and Planning within the Supreme Court. He referred to India's evolving legal-tech landscape and initiatives reshaping the courts including SUVAS, the Supreme Court's translation software that has enabled over 100,000 judgments in 18 regional languages, Automatic Speech Recognition (ASR) systems in Constitution Bench hearings for real-time transparency, and LegRAA, a legal research tool that aids without replacing judicial reasoning. 'These technologies are designed explicitly to support, not supplant, human judgment. It preserves the essential human element of jurisprudence, ensuring that final legal Page 6 of 13 interpretations remain firmly rooted in wisdom, compassion, and ethical discernment,' he said. Justice Kant called for building AI systems that reflect functional competence and moral clarity. 'I remain firmly convinced that any contemplation of AI must be guided by a deep moral compass. Shaping the future demands more than for an unwavering adherence to foundational values. Transparency, equity, responsibility, and respect for human dignity must not be afterthoughts, but the pillars upon which all technological advancement rests…Let this dialogue between technologists and jurists be not the end, but the beginning of a sustained collaboration, one where justice and technology walk hand in hand, with the citizen always at the centre.'
Time of India
2 hours ago
- Time of India
Delhi's e-rickshaw crisis: Numbering over 1.2L, why they should be city's last option for last-mile connect?
Representative AI image NEW DELHI: E-rickshaws are not a menace only the roads. They are also proving hazardous when parked or when their batteries are being recharged. Incidents related to charging are on the rise. In one incident reported last month, two people died after a fire broke out at a charging station in Shahdara. In another, a woman died and five people were injured when charging triggered a blaze in Farsh Bazar. While officially Delhi is estimated to have around 1.2 lakh registered e-rickshaws, the actual number that are operating could well be double the figure. Experts have serious concerns about safety and regulatory gaps in the e-rickshaw sector, particularly about the hazardous conditions of charging stations due to poor maintenance. Anil Chhikara, transport and vehicle expert, said some e-rickshaw drivers and owners have been observed to remove the original batteries and replace them with cheaper lead-acid batteries. "The latter batteries, when overcharged, can leak sulfuric acid — a highly corrosive substance that can cause severe burns or damage materials and people nearby," he said. The expert called for a thorough review and revision of the existing policies and regulations governing e-rickshaws, pointing out that in many cases, the charging points for e-rickshaws were neither secure nor compliant with safety standards. by Taboola by Taboola Sponsored Links Sponsored Links Promoted Links Promoted Links You May Like Free Solitaire. No Ads Ever. Play Solitaire Download Undo "There have been numerous instances when charging stations were found to be operating illegally, without any regulatory oversight or adherence to basic safety protocols," claimed Chhikara. The lack of dedicated charging stations in the city has also contributed to the increase in fires related to charging. SP Singh, vehicle expert, rued there were no concrete policies governing e-rickshaws, the first electric vehicles launched in the country. "There are proper charging stations for other electric vehicles, but not for e-rickshaws," noted Singh. "The places where these three-wheelers charge are not meant for that purpose and do not comply with safety standards. E-rickshaws get charged in large numbers at locations that have no awareness of or adherence to safety regulations." Singh also pointed out e-rickshaw drivers do not receive any training on safe charging practices or the precautions they should take while charging. "Delhi govt must establish separate fast charging stations across the city for e-rickshaws," he said. Lack of quality control in the e-rickshaw sector is also a major problem. Balraj Bhanot, former director, ARAI, and chairman, Transport Engineering Divisional Council, Bureau of Indian Standards, highlighted significant challenges faced by the e-vehicle manufacturing industry, particularly with regard to three-wheeler e-rickshaws. He noted that the industry was fragmented and decentralised across the country to meet local demands and often operated without adequate quality control. Firefighters said that when charging was carried out in residential buildings, the threat of accidents was higher. Atul Garg, director, Delhi Fire Service, reiterated that charging stations located in residential buildings always faced a significant risk of short circuits. He attributed the risks to use of poor-quality electrical equipment, improper or outdated wiring and overloading of the electricity setup. E-rickshaws have rapidly gained a notoriety for being unruly and reckless on the streets, frequently violating traffic rules and disrupting the smooth flow of traffic by clogging roads. Last year, Delhi Traffic Police recorded an increase of 47% in bookings related to violation of road laws by e-rickshaws. The number of offences rose sharply from 2,08,130 cases in 2023 to 3,05,959 in 2024. To curb the growth of illegal e-rickshaws, Delhi govt's transport department decided last year to scrap impounded and unregistered e-rickshaws within seven days if their owners did not apply for registration. The process previously allowed a leeway of 90 days for compliance.
