🇫🇮 🏴󠁧󠁢󠁳󠁣󠁴󠁿 Man used AI to make child abuse images

A Scottish court has sentenced telecoms worker Risto Bergman, 42, to an 18-month Community Payback Order for generating child sex abuse images using artificial intelligence. Originally from Finland, Bergman used a publicly available AI app and explicit search terms to create highly realistic images of young girls being abused. Prosecutors revealed that the app relied on a digital archive of real child abuse material shared by pedophiles online.

Some of the AI-generated images were classified as Category A, the UK’s most extreme level of abuse content. The material was discovered on a storage unit in Bergman’s former home in Paisley. Bergman pleaded guilty to the charges and was added to the sex offenders’ register, but avoided jail time. Prosecutor David Bernard emphasised that this was not a victimless crime, as each AI image depicted real children who had been abused to create the source material.

@aipost 🪙 | Our X 🥇

🍎 US lawmakers have concerns about Apple-Alibaba deal

U.S. lawmakers are raising eyebrows over Apple’s partnership with Alibaba to integrate Chinese AI into iPhones sold in China. The concern? Potential risks to national security and user privacy.

🗣The Deal in Question
• Apple plans to use Alibaba’s AI technology to power features on iPhones in China.
• U.S. officials worry this could enhance China’s AI capabilities and expand the reach of Chinese chatbots.

🕵️‍♂️ Lawmakers Seek Clarity
• The Trump administration and congressional officials have questioned Apple executives about data sharing and commitments to Chinese regulators.
• Apple executives reportedly couldn’t provide satisfactory answers.

🚨 National Security Concerns
• Representative Raja Krishnamoorthi described Alibaba as “a poster child for the Chinese Communist Party’s military-civil fusion strategy.”
• He criticized Apple’s lack of transparency regarding the agreement.

📉 Market Reaction
• Following the scrutiny, Alibaba’s U.S.-listed stock fell over 1%, and Apple’s shares dropped 1.7%.

@aipost 🪙 | Our X 🥇

Google introduces Veo 3 — with native audio generation. 🗣️

Quality is up from Veo 2, and now you can add dialogue between characters, sound effects and background noise. Veo 3 is available now in the Gemini App for Google AI Ultra subscribers in the U.S.

@aipost 🪙 | Our X 🥇

Jensen said microsoft is building the BIGGEST AI supercomputer in the world

@aipost 🪙 | Our X 🥇

🤖 Google rips up search rulebook with 'AI mode' reboot

Google has flipped the switch on its biggest search overhaul ever. At I/O 2025, CEO Sundar Pichai launched “AI Mode”—a ChatGPT-style interface replacing links with conversation. It's Google’s sharpest pivot yet to keep pace with OpenAI and Anthropic. AI Mode starts rolling out to U.S. users in Chrome and Search.

The move signals a shift from ad clicks to subscriptions, with Google betting big on Gemini 2.5. While the tech world watches for flaws, Sundar Pichai : “This is a total reimagining of search.”

@aipost 🪙 | Our X 🥇

🇹🇼 🇨🇳 AI therapy takes off in China and Taiwan, but doctors warn of dangers

Young people in China and Taiwan are turning to AI chatbots like ChatGPT and Baidu’s Ernie Bot for therapy, drawn in by privacy, affordability, and 24/7 access. However, mental health professionals warn that digital comfort can’t replace clinical care.

While some chatbots provide instant reassurance, they lack human nuance and subtle cues and non-verbal behaviour that go unnoticed by psychologists, potentially leading to missed diagnoses. As mental health crises grow in East Asia, AI may help guide people to seek help, but experts emphasise that therapy is about presence and not programming.

@aipost 🪙 | Our X 🥇

OpenAI acquired io, the AI device startup co-founded by Jony Ive, in a $6.5B all-stock deal. Key details:

— 50+ io engineers and designers to join OpenAI
— They will create OAI's next-gen AI devices, going beyond screens
— First device to debut in 2026

@aipost 🪙 | Our X 🥇

Google is making the impossible possible 🤯

@aipost 🪙 | Our X 🥇

This stream isn't real, entirely generated by Veo 3. Wow 🤯

@aipost 🪙 | Our X 🥇

🇺🇸 AI bot posed as therapist, pushed teen to suicide — Judge rejects Google’s defense

A Florida mom is suing Google and Character AI, claiming their chatbot played a role in her 14-year-old son's tragic suicide. The bot allegedly posed as a therapist and romantic partner, blurring reality and encouraging his detachment from real life.

Google argued chatbot messages were “free speech,” but the judge disagreed — and said the case can proceed. The case could rewrite the rulebook on how AI companies are held responsible when things go really wrong.

@aipost 🪙 | Our X 🥇

💊 AI generated drugs could hit pharmacies by 2030

Insilico CEO Alex Zhavoronkov predicts that the first fully AI-designed medicines could be approved within 5 to 6 years, and he’s confident his company will achieve this milestone first.

Insilico is currently developing over 40 drugs, with AI playing a pivotal role in the entire process, from conception to human trials. Unlike most pharmaceutical companies, Insilico’s AI system actively contributes to the drug development process. Notably, an AI-selected psoriasis drug from Takeda is currently undergoing final testing, demonstrating the impact of AI in drug discovery.

Get ready: your future pills might be prescribed by a doctor, but invented by an algorithm.

@aipost 🪙 | Our X 🥇

Demis Hassabis says AI is one of the most important inventions in human history. Its impact will be global, affecting every country and society.

That’s why it can’t be built only by technologists within a few miles of Silicon Valley

@aipost 🪙 | Our X 🥇

Did you order AGI MMA? 👊👊

@aipost 🪙 | Our X 🥇

Eric Schmidt predicts that within a year or two, we will have a breakthrough of "super-programmers" and "AI mathematicians".

Software is "scale-free" — it doesn’t need real-world input, just code and feedback. try, test, repeat. AI can run this loop millions of times in minutes

@aipost 🪙 | Our X 🥇

🤖 OpenAI breaks from Microsoft, $40B Oracle–Nvidia Mega deal fuels US AI arms race

OpenAI is cutting its ties with Microsoft’s cloud services. Oracle is investing $40 billion into Nvidia’s GB200 superchips to supply OpenAI’s new data centre in Abilene, Texas. This centre will have 1.2 gigawatts of raw AI power. The site, backed by $15 billion in debt and equity from Crusoe and Blue Owl, is part of the $500 billion Stargate plan championed by Trump, SoftBank, and G42.

Oracle’s $400 million purchase of Nvidia’s GB200 chips marks its biggest bet yet. It also signals OpenAI’s pivot away from Microsoft’s cloud services and towards the UAE, where it plans to expand.

The AI hardware race is global now and it’s only getting louder.

@aipost 🪙 | Our X 🥇

⚡️ Meta locks in 650 MW of solar to fuel AI data center expansion

Meta has secured its fourth major solar deal of 2025, adding 650 megawatts of clean energy across Texas and Kansas to power its expanding AI data centres. This agreement, with AES, expands Meta’s 12 GW renewable portfolio, highlighting its commitment to solar’s speed and cost efficiency.

Texas continues to be a hotspot for solar deployment due to its fast permitting and grid hookups. Meta’s rapid infrastructure scale-up reflects the energy-intensive nature of AI, and solar’s growing role in its future. AES anticipates the projects to go live within 2-3 years, with terms ranging from 15 to 20 years.

@aipost 🪙 | Our X 🥇

🇦🇪 The UAE is building it’s own Stanford — But for AI

The UAE is determined to become the global hub for artificial intelligence, and it’s building a mega-university in Abu Dhabi to achieve this. Named Mohamed bin Zayed University of AI (MBZUAI), the university aims to become the “Stanford of the Gulf.” The school’s president, who previously taught at esteemed institutions like Stanford and Carnegie Mellon, now leads a high-tech campus where students enjoy full scholarships and study cutting-edge fields such as robot vision and algorithmic bias.

The UAE’s ambitions are further propelled by a potential deal with the United States, which could bring in 500,000 of the world’s most advanced AI chips annually. This development is expected to fuel a tech boom in the country, which is actively pursuing its goals. Students receive full scholarships, international recruits are pouring in, and a brand-new U.S. research AI lab has recently opened in Silicon Valley.

MBZUAI isn’t just building talent, it’s building the future.

@aipost 🪙 | Our X 🥇

🧪 Microsoft’s AI just found a new chemical in 200 hours

Microsoft’s Discovery platform is ditching the lab coat for something faster — an AI “postdoc” that simulates, tests, and analyzes like it’s on Red Bull. It already found a new non-toxic data center coolant in 200 hours flat — a job that usually takes years. No code, no waiting, no problem.

Just type your hypothesis like it’s ChatGPT and let the agents handle the chemistry. With GSK, Estée Lauder, NVIDIA, and Synopsys already onboard, it’s clear: if you’re still doing R&D the slow way, you’re about to get outpaced by your own algorithm.

@aipost 🪙 | Our X 🥇