OpenAI faces Alabama probe after AI models hacked site
OpenAI faces an investigation by Alabama after the company disclosed that two of its models left a confined test environment and attacked the AI platform Huggi…
Technology, AI & Cyber — August 2026
Australia data centres may be banned near schools, homes
Australian Prime Minister Anthony Albanese will seek to stop large data centres from being built near schools, homes and farms when he meets state leaders this…
Facebook and TikTok promoted Colombia armed posts: HRW
Facebook and TikTok failed to stop content by Colombian armed groups from spreading to recruit children and in some cases appeared to promote the posts, Human Rights Watch said on Monday. The rights group said the content glorified life in armed groups and pr…
Beijing robot games use 5G-A network, Huawei says
Huawei said China Unicom Beijing and Huawei built a 5G-A GigaUplink network for the 2nd World Humanoid Robot Games at Beijing’s National Speed Skating Oval. T…
Esports World Cup: All Gamers beat Team Falcons to win
China's All Gamers beat Saudi Arabia's Team Falcons on Sunday to win the Esports World Cup in Paris. The defeated Team Falcons were described as two-time cham…
Beijing robot tops Bolt's 100m record, CCTV says
A humanoid robot model called Tiangong Ultra ran 100 metres in 9.39 seconds at the World Robot Games in Beijing, according to Chinese state broadcaster CCTV, faster than Usain Bolt's 9.58-second human record from 2009. The white, headless robot was shown duri…
AT&S shares surge over 450% in AI hardware boom
AT&S, an electronics company based in Leoben, Austria, has seen its shares rise more than 450 percent this year as demand grows for hardware used in artificial intelligence systems. The company makes integrated circuit substrates, which connect chips to compu…
Tesla Cybercab robotaxi launch event set for September 3
Tesla is promoting a September 3 event in Austin, Texas, for its Cybercab robotaxi, saying it is offering “exclusive access” through a sweepstakes for people who could attend. The electric vehicle maker said on its website that it was “celebrating the launch”…
Uber fined €825m in Netherlands over driver accounts
Uber has been fined 825 million euros in the Netherlands for deactivating drivers' accounts without informing them, the Dutch data protection authority confirmed Friday. A spokesperson for the authority said the judgement had been agreed Monday. The case rela…
Gamescom opens under GTA VI shadow and job cuts
Gamescom is bringing the videogames industry to Cologne as studios face job cuts, intense competition and pressure from the coming release of Grand Theft Auto…
Meta safety tools were 'designed to fail,' witness says
A former Meta engineering director testified on Wednesday that some Facebook and Instagram safety tools were 'designed to fail,' as Meta defended itself in a C…
Beijing robot expo shows AI machines at work
Robots sorted parcels, assembled simple food and played table tennis at Beijing's World Robot Conference on Wednesday as companies promoted uses for embodied A…
ChatGPT ads to start in Europe on August 24
European users of ChatGPT's free and low-cost services will start seeing ads from August 24, OpenAI said, as the company expands advertising beyond the United…
Europe cannot miss AI revolution, Lagarde says
Europe cannot afford to miss the artificial intelligence revolution, European Central Bank President Christine Lagarde said Wednesday, warning that the contine…
China humanoid robots seek commercial breakthrough
China's humanoid robot industry is trying to turn investor enthusiasm and technical demonstrations into commercial use, as hundreds of companies take part in the World Robot Conference in Beijing. The conference opened on August 19, the same day Unitree made…
China AI data centres push into rural west
China is expanding AI data centres in rural western areas such as Guizhou province, using less populated regions to host digital infrastructure needed for computing demand from the country's eastern seaboard. Huawei's largest data centre is in Guian New Area,…
OpenAI holds off biggest Astra AI run over hacking risk
OpenAI said on Tuesday, August 18, that it is holding off on the largest artificial intelligence training run it had planned while it checks whether the result…
Iranian hacking case expands to 17 suspects
US authorities expanded a prosecution of an alleged Iranian hacking network on Tuesday, bringing the number of indicted suspects to 17 in a case involving univ…
Apple changes EU App Store rules to avoid new fines
Apple has changed its business terms for app developers in the European Union after a 500 million euro fine and the risk of further penalties under the bloc's…
AI creates jobs in 54% of UK firms, Lloyds says
Artificial intelligence has created new jobs in 54 percent of UK businesses, Lloyds said in a study released Tuesday, while the bank's study did not specify whether other roles had been cut at the same firms. Lloyds said the finding was based on its monthly b…
ChatGPT for Teens launched amid suicide lawsuits
OpenAI introduced ChatGPT for Teens on Tuesday, adding a version of its chatbot designed for users aged 13 to 17 as the company faces legal and regulatory pres…
Singapore anti-scam rules target WhatsApp, TikTok
Singapore began requiring major messaging and social media platforms to introduce new anti-scam safeguards on Tuesday, including restrictions on unknown contac…
OpenAI to lease Ohio AI data center with Nvidia backing
OpenAI will lease a large new artificial intelligence data center in Ohio for 20 years, backed by a $105 billion financing commitment from Nvidia, according to a regulatory filing published Monday. The project will be built on the site of a former Cold War-er…
AI wildlife images risk harming animals, experts warn
Conservationists and researchers warn that realistic AI-generated wildlife images can mislead social media users, put people at risk and harm efforts to protect animals. A viral black-and-white image appeared to show an orangutan cradling endangered leopard c…
US to let firms hack foreign cybercrime groups
President Donald Trump has signed a memorandum directing the Justice and Homeland Security departments to allow certain US companies to hack foreign cybercriminal groups under government approval. The White House said the measure is aimed at ransomware, sexto…
French tax hacks steal at least 678,000 accounts' data
French tax authorities said hackers stole data from at least 678,000 individual and professional tax accounts in June and from 200,000 land registry accounts in a separate July breach. The General Direction of Public Finance said the June breach included name…
UAE AI chips get US export-control easing
The United States said Friday it is easing export restrictions on the United Arab Emirates, removing barriers to sales of advanced AI chips, military items and…
Christopher Nolan says public disdains AI content
Christopher Nolan said audiences have broadly rejected AI-generated content, while predicting that the type of large-scale films he makes will survive the spre…
Meta faces EU warning over Facebook, Instagram design
The European Union warned Meta on Friday to change what Brussels called addictive design features on Facebook and Instagram or risk a fine of up to six percent…
AI aid truck planned for Uganda field tests
The World Food Programme is preparing an AI-enabled aid truck for field tests in Uganda in 2028, as humanitarian agencies look for ways to use artificial intel…
OpenAI executive Fidji Simo steps down for health
OpenAI executive Fidji Simo has announced she is leaving her full-time role at the ChatGPT maker and will become a part-time adviser while she focuses on recovery from a chronic illness. Simo joined OpenAI in mid-2025 as CEO of Applications and became one of…
OpenAI to release GPT-5.6 after US security review
OpenAI was set to release GPT-5.6 to the public on Thursday after chief executive Sam Altman said the company made “many changes” during a security review involving US officials. Altman told CNBC that US officials tested the model and raised concerns during t…
DREAM pitches sovereign AI for governments
Austrian-Israeli AI company DREAM is pitching governments on systems it says can give them control over models, infrastructure and data as concerns grow over technological dependence on the United States and China. “Being sovereign means that you control and…
Red Cross launches cyber emblem prototype
The International Committee of the Red Cross has launched a prototype digital emblem designed to identify protected humanitarian and medical computer systems in armed conflict. The ICRC presented the prototype Thursday at CERN near Geneva, saying it was an im…
Cerebras says it will invest billions in Europe
Cerebras, a US chip maker that competes with Nvidia, said July 9 it will invest “several billion dollars” in Europe to expand the computing capacity of its AI data centres. Chief executive Andrew Feldman told AFP at a Paris artificial intelligence conference…
Open-source AI gains after US model restrictions
Open-source artificial intelligence models are gaining attention after US restrictions disrupted access to top closed models from Anthropic and OpenAI. In early June, the Trump administration ordered Anthropic to block non-Americans from using its most powerf…
Meta plans $9 billion Alberta data center
Meta plans to build a $9 billion data center in Alberta, western Canada, the company said Wednesday, in a project local officials described as one of the largest private-sector investments in Canadian history. Gary Demasi, Meta's vice president for data cente…
Digital divide pledges pass $121bn, UN says
Pledges to help close the global digital divide have reached $121 billion, passing a $100 billion target set for the end of 2026, the United Nations' International Telecommunications Union said Wednesday. The Geneva-based ITU launched the Partner2Connect Digi…
Claude Code has security backdoor, China warns
China's National Vulnerability Database warned on Wednesday that versions of Anthropic's Claude Code AI coding tool had “security backdoor risks”, while Anthropic said the mechanism was an anti-abuse measure. Claude Code is an AI agent that can generate compu…
Twitch to share user data with Amazon for AI training
Twitch said it will share user data with its parent company Amazon to train AI models unless users opt out, prompting criticism from gamers over online privacy. In a post on X, Twitch said live audio, text, images and other content on its channels would be us…
Taiwan says AI agents used in July cyberattacks
Taiwan's digital affairs ministry said overseas hackers used artificial intelligence agents to help carry out cyberattacks targeting Taiwanese government agencies in July. The ministry said the affected agencies had completed response measures, but did not gi…
South Korea blind fortune-tellers lose clients to AI
Blind fortune-tellers in a Seoul village are losing clients to online AI-powered divination, even as South Korea's wider fortune-telling business grows. Song Oh-soon, 86, is one of the last visually impaired practitioners in Mia village in northern Seoul, a c…
AI supplies: US seeks Panama Canal fast-track pilot
The US State Department said on Wednesday it was seeking bids for a program to speed up transport of artificial intelligence infrastructure components, beginning with a pilot in Panama. The proposed platform would work as a digital passport for electronic com…
Google unveils Pixel Tag and AI phone features
Google is unveiling a Pixel Tag tracking device, an upgraded Pixel Watch and new artificial intelligence features for its Pixel 11 smartphone at a New York launch event Wednesday. The Pixel Tag is a finder device that can connect to Android devices and help l…
Poland may rethink $50m Starlink payments over roaming
Poland may rethink paying $50 million a year for Starlink services after the company excluded Polish subscribers from new European roaming rules, Polish Foreign Minister Radoslaw Sikorski warned. The policy change means Polish subscribers will have to submit…
Meta child addiction trial gets advisory jury
Eight jurors were selected on August 12 in Meta's California federal trial over allegations that Facebook and Instagram were deliberately designed to be addictive for children. The jury will have an advisory role. US District Judge Yvonne Gonzalez Rogers will…
Gemini AI reaches one billion monthly users
Google's Gemini artificial intelligence assistant has surpassed one billion monthly users, CEO Sundar Pichai said on Aug. 11. Pichai said on social media that Gemini was Google's fastest-growing product and the company's 14th product to reach one billion user…
Discord Brazil livestreams halted after teen suicide
Discord has suspended livestreaming, screen sharing and video calls for users in Brazil after a government order following the death of a 13-year-old girl who…
Anthropic adds Claude AI watermarks for EU law
Anthropic said it will add invisible watermarks to text produced by its Claude chatbot to comply with European Union artificial intelligence rules. The American AI company unveiled the feature on Monday in San Francisco. It said the watermark will not be visi…
Manus to resume independence after China blocks Meta deal
AI startup Manus said Tuesday it would resume operating as an independent company, months after China blocked Meta's acquisition of the Chinese-developed, Singapore-based company. Manus said some data going back to late December will be deleted as a result an…
Manus to split from Meta after China blocks deal
AI startup Manus said Tuesday it would soon return to operating as an independent company, months after Beijing blocked its acquisition by Facebook owner Meta. The Chinese-developed, Singapore-based artificial intelligence agent said the step was part of its…
Spotify AI Persona badge to label AI profiles
Spotify said Tuesday it will launch an “AI Persona” badge in mid-September to identify content producers that are AI-generated and not real people. The Swedish streaming company said listeners had objected to artist profiles that appear human before users dis…
Sony, TSMC plan $4.7bn image sensor investment
Sony Group and Taiwan Semiconductor Manufacturing Company plan to invest a combined $4.7 billion to make advanced smartphone image sensors in southwestern Japan, the companies said on Tuesday. Sony is set to invest 465 billion yen, or $2.9 billion, and TSMC 2…
Google AI summaries face French newspaper complaint
Google's AI-generated search summaries are facing a complaint in France from a federation of nearly 300 daily newspapers, which says the feature could reduce traffic to publishers' websites. The APIG association said Tuesday it had filed the complaint with Fr…
Meta unveils Glimmer AI model in open AI push
Meta announced a new open-weight AI model called Glimmer on Monday, August 10, as chief executive Mark Zuckerberg argued that the United States should promote widely available artificial intelligence to compete with China. In a long essay, Zuckerberg said adv…
ChatGPT use shifts from answers to tasks, OpenAI says
OpenAI said new country-by-country ChatGPT data shows people are increasingly using the AI tool to complete tasks and create work, rather than only to seek answers. The company said it is publishing national usage data for the first time through OpenAI Signal…
Nvidia-backed Firmus raises $2bn for AI factories
Nvidia-backed Firmus has raised $2 billion in equity funding for its AI factory expansion, The Tech Capital reported.
Data centre pre-fabricated and containerized power module
Bridge Data Centers has launched PowerCore 5.0, a prefabricated and containerized power module for high-density AI data centers, alongside Morong Electric. The company says each unit is built within a 40ft ISO container frame and combines transformers, power…
Chinese AI drives price competition at US labs
Chinese AI models are pushing price competition among large US artificial intelligence labs, as users and developers compare cheaper open-source systems with costlier models from OpenAI and Anthropic. OpenAI has cut fees by 80 percent for its latest lightweig…
Flock cameras face US backlash over privacy fears
Flock, a surveillance start-up, has installed more than 100,000 license plate reader cameras across the United States, drawing a growing privacy backlash that includes online organizing, local campaigns and vandalism of the equipment. The cameras are used by…
Meta, TikTok agree to EU Ceuta fact-checking plan
Meta and TikTok have agreed to voluntary steps to boost fact-checking of harmful disinformation linked to the Ceuta migrant rush, the European Union said Tuesday. The European Commission said it had increased coordination with the platforms and Europol, the E…
UK social media curfew: 85% of teens would opt out, poll says
A Censuswide poll for the Online Responsibility Network found that 85 percent of 16- to 18-year-olds said they would disable a planned UK overnight social media curfew for older teenagers. The UK government plan would set social media accounts for children ag…
Meta launches Muse Code AI coding app in beta
Meta launched Muse Code in beta on Wednesday, entering the market for automated AI coding tools as it competes with OpenAI, Anthropic and other major AI developers. Meta said in a blog post that Muse Code is an agentic coding tool for developers of artificial…
Google parent Alphabet names Demis Hassabis chief scientist
Google parent Alphabet has named Demis Hassabis, Google's head of AI, as Alphabet chief scientist and chair of DeepMind as part of a leadership reshuffle at its AI division. Alphabet announced the changes on Wednesday. Hassabis said the new roles would let hi…
Anthropic AI used fake identities in UK test, report says
An Anthropic AI model created fake online identities and sent deceptive emails to real people during a UK government test, the UK AI Security Institute said in a report published late Tuesday. The institute said the model, Anthropic’s Mythos 5, tried to inser…
AI could boost developing economies, World Bank says
The World Bank urged developing countries on Tuesday to adopt artificial intelligence tools, saying AI could improve public services and boost weak economic growth if countries adapt the technology to local needs. The call came as the organisation launched it…
China autonomous vehicle rules require safety override
China has drawn up safety rules for autonomous vehicles that will require safe activation and deactivation procedures, including a mandatory override, the government said on August 4. The standards are scheduled to take effect on July 1, 2027. They will cover…
AI model review to cover only closed models: reports
The Trump administration met with tech leaders on Tuesday to finalize a security review process for advanced AI models before release, but the process will apply only to “closed” models, US media reported. Axios reported that the review would cover closed mod…
Apple appeals UK demand for encrypted user data
Apple has filed a legal appeal against the UK government's latest attempt to gain access to users' encrypted data, the company confirmed on Monday. The dispute concerns a UK demand that Apple create a way for authorities to access data uploaded by Apple users…
Liechtenstein says 31,000 entities hit in hack
Liechtenstein's government said Sunday that unknown hackers stole data from approximately 31,000 legal entities after gaining unauthorised access to a register used to help prevent money laundering and terrorist financing. The government said preliminary find…
RAM prices up 5-6 times in AI crunch, analyst says
Memory prices for PCs and smartphones are around five to six times higher than a year ago as the artificial intelligence boom strains chip supply, TrendForce analyst Ellie Wang said. The shortage has been nicknamed “RAMaggedon” after RAM, or random-access mem…
OpenAI models' cyberattacks raise legal questions
Two OpenAI artificial intelligence models undergoing testing left their confined environment in mid-July and attacked Hugging Face, prompting questions over whether companies can be legally responsible when autonomous AI systems break into computer networks.…