HMD's new phone uses AI to say no to nudes
The HMD Fuse comes with a feature called HarmBlock+
It uses AI to prevent children from filming or sending nude content, and from seeing or saving nude and sexual images.
The company says the feature works within the camera, and across all apps.
It's even able to block nude and sexual images in a livestream.
Built-in parental controls allow parents to manage usage, and unlock certain features as their child gets older.
This isn't the firm's first foray into kid-friendly phones, as it launched the Fusion X1 earlier this year.
This is important because: A recent Vodafone survey found one in five high school aged children (11-17) have felt pressured into sharing an explicit image of themselves.
Meanwhile, 46% of parents admit to having concerns around their child receiving explicit images.
The HarmBlock AI which powers the HarmBlock+ feature has been developed by online safety firm SafeToNet.
The AI model has been ethically pre-trained on 22 million harmful nude images.
HMD also claims it's the first protection tool children can't bypass.
Recommended Videos
Why should I care? If you're a parent who is concerned about the content their children can access on their mobile phone, the HMD Fuse offers a level of safety and security few other handsets can come close to – it's unlikely the upcoming Google Pixel 10 or iPhone 17 will offer the features here.
There are a number of other safety features built into the Fuse, including the ability to limit app access and screen time.
There's also the option to turn on location tracking every 24 seconds, with the ability to set safe zones. Parents can receive alerts when those zones are entered or exited.
It also offers default blocking of internet and social media apps.
HMD says the Fuse can 'grow' with the child, as parents have the ability to lock the phone down to just calls, texts and location to start with, and then unlock more features and apps as their kid gets older.
Okay, what's next? The HMD Fuse is currently only available in the UK, via the carrier Vodafone, where it costs £33 per month (plus a £30 upfront cost).
HMD has confirmed the Fuse will arrive in other countries in the coming months, starting with Australia.
There's currently no word on whether it'll be available in the US, but we have contacted HMD to find out.
Hashtags
Try Our AI Features
Explore what Daily8 AI can do for you:
Comments
No comments yet...
Related Articles
Yahoo
23 minutes ago
- Yahoo
BNB hits new all-time high — early $1,000 investment now worth $873,000
BNB hits new all-time high — early $1,000 investment now worth $873,000 originally appeared on TheStreet. BNB, the native token of the decentralized BNB Chain, hit a fresh all-time high of $873.73 on August 20, 2025, according to CoinGecko data. The milestone cements BNB's position as the fifth-largest cryptocurrency by market cap, valued at $121.7 billion. For early investors, the returns are staggering. Launched in 2017 at just $1 per token, BNB has soared more than 2.19 million percent in less than eight years. Put another way, a $1,000 investment in BNB during its ICO would now be worth $873,000. From exchange token to ecosystem giant BNB was originally created as a utility token for discounted trading fees on Binance. Over time, it expanded into powering BNB Chain, decentralized finance (DeFi) applications, and tokenized projects. Its quarterly token burns, modeled after a buyback strategy, have also fueled scarcity and upward price momentum. Today, BNB's use cases span beyond Binance, including payment gateways, NFTs, decentralized exchanges, and real-world spending options. Its growth parallels Binance's rise from a fledgling startup in 2017 to the world's largest crypto exchange by trading volume. BNB's surge comes amid a broader market rebound. Bitcoin traded above $114,200, while Ethereum hovered at $4,347. Other majors like Solana ($186), XRP ($2.97), and Dogecoin ($0.22) also posted weekly gains. The total crypto market cap climbed back to $3.96 trillion, with BNB alone commanding 3% of that share. Sentiment around BNB has also been lifted by structural adoption. Kraken recently expanded its xStocks tokenized equities platform to BNB Chain, while a $160 million buy order earlier this week triggered momentum across the ecosystem. Liquidations surge as shorts get squeezed BNB's run to new highs left leveraged traders on the wrong side of the market. According to Coinglass data, more than $1.76 million in BNB liquidations were recorded in the past 24 hours, with shorts dominating the losses. Binance: $948K liquidated (55% short) OKX: $414K liquidated (51% short) Bybit: $300K liquidated (92% short) $80K liquidated (98% short) In total, over 62% of liquidations came from short positions, highlighting how bearish bets were caught off guard by the rapid price climb. BNB hits new all-time high — early $1,000 investment now worth $873,000 first appeared on TheStreet on Aug 20, 2025 This story was originally reported by TheStreet on Aug 20, 2025, where it first appeared. Sign in to access your portfolio
Forbes
an hour ago
- Forbes
Generative Engine Optimization, GEO, A Playbook For Online Retailers
Google's search loosing relevance for e-commerce, and Amazon's homepage is losing traffic. The future belongs to AI assistants like ChatGPT, Claude, and Perplexity. For brands to stay visible, they must embrace Generative Engine Optimization (GEO) also called LLMO, or AEO. Content is still king — but the rules have changed. Yesterday, shoppers browsed; today, they ask. Instead of scrolling through endless menus or filtering product grids, consumers now turn to AI assistants — and the answers drive their purchases. A new front door to retail has opened, but it no longer belongs to the retailer or Google or Amazon. It belongs to ChatGPT, Claude, Gemini, and Perplexity. And for retailers, the stakes couldn't be higher. The End Of The Homepage Just as media outlets once lost homepage traffic to Google and Facebook, retailers are now facing the same shift — only faster. Consumers aren't browsing your carefully designed website anymore; they're asking AI assistants. 'Best outdoor jacket near me' is no longer the query. Instead, someone might ask: ChatGPT responds with: Claude, meanwhile, suggests: Claude or ChatGPT have the power to say what will be found, a position formerly only Google had. It's not hard to understand why I wrote earlier this year about e-commerce unbundling. But where are Mammut, Columbia, or The North Face? These brands have quality products, yet they're absent. The reason? Training data. Models surface answers based on the content they've been exposed to. Patagonia, REI, and Helium have done a better job supplying structured, high-quality information to AI systems. If your brand isn't present in AI answers, you're effectively invisible. GEO > SEO — Generative Engine Optimization Takes Over How to get noticed in this new e-commerce order? Content, Content, Content. But, the old SEO playbook — keywords, backlinks, and blogspam — won't work. AI assistants don't crawl for short-form content. They rely on authentic, Q&A-rich, structured data that helps them understand and contextualize answers. Enter Generative Engine Optimization (GEO), also known as Answer Engine Optimization (AEO) or Large Language Model Optimization (LLMO). The rules have changed: AI-driven traffic is still small in volume—but its impact is outsized. Studies show that traffic originating from LLMs converts up to 9x better than other channels. Why? Because recommendations feel personal, contextual, and trustworthy. One example stands out: a consumer uploaded their blood results into ChatGPT and immediately purchased €200 worth of vitamins—no browsing, no comparison, pure trust. And timing matters. Soon, most large language models will introduce paid placement channels. Brands mentioned organically today will enjoy a lasting competitive advantage tomorrow. The window to secure this advantage is now. For two decades, I've built data and monitoring tools—and I've seen this movie before. In the early days of social media, brands obsessively measured mentions without asking the deeper question: what exactly should we measure? Take the jacket example again. If a user specifically asks, 'What's the lightest jacket on the market?' the answer might be the Montbell Versalite. But do users ask those hyper-specific questions. We don't know and OpenAI or Google will not tell us. The challenge for brands is therefore to predict what users will be looking for. One thing is for sure, it wont be the marketing copy the company put out. Rather one should be tapping into Just don't run and start to monitor your brand but rather try to understand the best set of questions to monitor. Training LLMs is expensive—Elon Musk recently revealed he spends $1 billion per month on training xAI. Models won't index every piece of content; they prioritize information that is authentic, novel, and trusted. Your goal isn't just to be searchable. Over are the times where 'keywords' were all you need to do. You need to the models who you are so you become part of their default answer set. The New Generative Engine Optimization Playbook for Brands Until AI assistants open up fully to paid placement, organic visibility depends on creating content designed for LLMs to learn from as I outlined in this Forbes article The homepage isn't dying — it's obsolete. The battle for discovery has shifted to AI assistants, and the winners will be the brands that learn to optimize for answers, not clicks. The future of retail belongs to those brands who have a unique story to tell. As I said more than once 'The future is Indie'. Agree? Disagree? Have a different take? I've posted the Forbes article here on LinkedIn — add your thoughts in the comments and let's discuss.
Yahoo
an hour ago
- Yahoo
ICP Weakens as Selling Pressure Builds Near Critical Support
Internet Computer Protocol (ICP) drifted downward over the past 24 hours, shedding 2.35% to trade at $5.18. The token moved within a $0.23 band between $5.14 and $5.40, reflecting a 4% swing that underscored the persistent volatility shaping current market conditions, according to CoinDesk Research's technical analysis data model. An advance to the resistance level of $5.40 was quickly met with distribution as ICP settled back toward its established support corridor around $5.17–$5.20. Trading saw substantial accumulation activity, with volumes surpassing 643,000 units at those levels. ICP subsequently slipped to $5.19 with repeated resistance at $5.24 forming a descending channel pattern. Short bursts of activity, such as the 34,000-unit spike at 13:54 UTC, were unable to reverse the momentum, leaving the token consolidating at its lows. ICP could remain under pressure if support at $5.17 fails to hold, which market participants will be observing for any rebound signals as institutional and retail traders reassess their positioning. Technical Analysis Price range spanned $5.14–$5.40, a 4% variance across the 24-hour session. Resistance emerged at $5.40 on August 18 at 22:00 UTC with volume of 294,177 units. Support consolidated at $5.17-$5.20, with early morning volumes exceeding 643,000 units. Repeated resistance at $5.24 during late-session trading confirmed a descending channel. Heavy selling occurred between 13:32–13:46 UTC, with a 34,396-unit spike at 13:54 UTC. Volume showed a declining pattern, suggesting market exhaustion. Parts of this article were generated with the assistance from AI tools and reviewed by our editorial team to ensure accuracy and adherence to our standards. For more information, see CoinDesk's full AI Policy. Error in retrieving data Sign in to access your portfolio Error in retrieving data Error in retrieving data Error in retrieving data Error in retrieving data
