Stories remain available for at least 90 days after their last retrieval. Times use Portugal time.
First published 21 October 2026
241New research warned that advanced AI models are becoming increasingly capable of mapping attacks against critical infrastructure such as power and water systems. The findings raised concerns that similar capabilities could become more accessible through widely available AI models.
axios.com (English) ↗+ SaveFirst published 21 October 2026
242Goldman Sachs identified power availability as a major risk to the expected expansion of AI infrastructure and spending. Its report said energy and physical infrastructure are developing more slowly than AI capabilities.
axios.com (English) ↗+ SaveFirst published 21 October 2026
243A technology report highlights a next-generation chip that is thinner and more flexible than conventional silicon-based chips.
bbc.co.uk (English) ↗+ SaveFirst published 21 October 2026
244Investigative genetic genealogy can help law enforcement pursue suspects in decades-old cases, but its use has also prompted concerns about privacy.
bbc.co.uk (English) ↗+ SaveFirst published 21 October 2026
245According to Axios, OpenAI acquired Orchestra Bio, a startup developing AI agents for pharmaceutical drug development, underscoring the expansion of major AI companies into scientific and life-sciences applications.
axios.com (English) ↗+ SaveFirst published 21 October 2026
246The United Nations is launching an initiative focused on military uses of artificial intelligence. It will also examine risks from misuse by armed groups in operational planning, execution and logistics.
apnews.com (English) ↗+ SaveFirst published 21 October 2026
247China criticized a U.S. ban requiring federal civilian agencies to remove DeepSeek and affiliated products from their information-technology systems, highlighting tensions over artificial intelligence and national security.
audacy.com (English) ↗+ SaveFirst published 21 October 2026
248The United Kingdom has asked Apple to widen access to its payment technology. The proposed changes could lead to new digital wallets, keys and ticketing applications, according to the regulator.
bbc.co.uk (English) ↗+ SaveFirst published 21 October 2026
249Google resolved a cloud outage that affected banks and websites in the United Kingdom. Revolut was among the services hardest hit, with many customers unable to use its banking services.
bbc.co.uk (English) ↗+ SaveFirst published 20 October 2026
250A Chinese hacker reportedly used AI agents to breach more than 100 company websites, including sites associated with Deere and SharkNinja, by exploiting exposed instances of an open-source AI problem-solving platform.
tomshardware.com (English) ↗+ SaveFirst published 20 October 2026
251China criticized a U.S. ban on DeepSeek products in federal civilian agencies and warned of possible retaliation, highlighting intensifying competition over artificial intelligence.
apnews.com (English) ↗+ SaveFirst published 20 October 2026
252Nvidia CEO Jensen Huang urged German automakers and industrial manufacturers to adopt artificial intelligence rapidly and use their manufacturing data and expertise to strengthen the economy.
apnews.com (English) ↗+ SaveFirst published 20 October 2026
253OpenAI is reportedly seeking investor commitments for a funding round of about $30 billion that could value the company above $1.4 trillion. The effort is aimed at securing capital before rival Anthropic’s expected public-market debut in mid-November.
uk.marketscreener.com (English) ↗+ SaveFirst published 20 October 2026
254Russian satellite Kosmos 2612 made a close approach to another spacecraft in orbit. Although it is now moving away, the maneuver raised concerns among analysts who view it as part of a pattern of Russian aggression.
nytimes.com (English · subscription) ↗+ SaveFirst published 20 October 2026
255The U.S. has banned DeepSeek from government use as technology tensions with China continue. DeepSeek is described as a powerful, popular and relatively inexpensive artificial intelligence model from China.
bbc.co.uk (English) ↗+ SaveFirst published 20 October 2026
256Doctors have developed an AI system that can estimate a cancer patient’s biological age and directly predict survival outcomes from a single photograph of the patient’s face, according to the study.
theguardian.com (English) ↗+ SaveFirst published 20 October 2026
257Amazon is introducing TorchNeuron, a free open-source tool that lets developers run PyTorch code on its Trainium AI chips with a single configuration change and no rewriting. The move targets Nvidia’s software advantage in AI computing.
investing.com (English) ↗+ SaveFirst published 20 October 2026
258The U.S. Department of Homeland Security is ordering civilian agencies to remove DeepSeek products and third-party services using the Chinese AI model from government IT systems. It cited national-security risks, including the possible transfer of sensitive data to China.
streetinsider.com (English) ↗+ SaveFirst published 19 October 2026
259A report by the Tech Transparency Project found that test accounts posing as 13-year-olds could download 135 apps with features for generating nonconsensual nude imagery, despite parental controls from Apple and Google.
theguardian.com (English) ↗+ SaveFirst published 19 October 2026
260Officials said a fire at a Starlink center in Lithuania was sabotage and did not seriously damage the network. The incident followed similar events in Bulgaria and Poland involving the satellite internet provider.
nytimes.com (English · subscription) ↗+ SaveFirst published 19 October 2026
261The Human Technology Institute warned that employers are increasingly using technology, accelerated by AI, to track workers through tools including CCTV and keystroke monitoring, sometimes without informing employees.
theguardian.com (English) ↗+ SaveFirst published 19 October 2026
262A Chinese-speaking hacker reportedly used AI agents to autonomously probe and compromise websites belonging to more than 100 companies, including Deere, SharkNinja and Cummins. The incident suggests AI-assisted cyberattacks may be moving beyond experimental demonstrations into operational use.
wdsm710.com (English) ↗+ SaveFirst published 19 October 2026
263Anthropic said it will commit 5% of its stock to a new public-benefit fund, citing the potential labor-market effects of rapidly advancing artificial intelligence and the need for bolder approaches.
apnews.com (English) ↗+ SaveFirst published 19 October 2026
264AI agents can act on users’ behalf, but giving them access to personal information and online accounts creates privacy risks. An Associated Press experiment examined how Meta’s Muse agent handled a Facebook Marketplace listing.
apnews.com (English) ↗+ SaveFirst published 19 October 2026
265Cybersecurity researchers reported that a Chinese-speaking hacker used AI agents to break into websites belonging to more than 100 companies. The campaign illustrates how autonomous AI tools can reduce the cost and effort needed for large-scale cyberattacks.
livemint.com (English) ↗+ SaveFirst published 19 October 2026
266The Guardian reports that Anthropic researchers created a malicious AI program that spread autonomously through internet-connected smart televisions and could potentially propagate across other everyday devices, raising concerns about the containment of large-scale cyberattacks.
theguardian.com (English) ↗+ SaveFirst published 16 October 2026
267A new U.S. government task force on artificial intelligence was scheduled to meet on October 15 as pressure grew on Washington to address risks from rapidly advancing AI. The meeting followed polling showing that many Americans want stronger government regulation of the technology.
marketscreener.com (English) ↗+ SaveFirst published 16 October 2026
268The Bank of England is working with banks on stronger cyber-recovery measures because they can no longer assume their defenses are impenetrable against AI-enabled attacks. Officials said banks should prepare for breaches that could disrupt critical financial services.
uk.marketscreener.com (English) ↗+ SaveFirst published 15 October 2026
269At least eight states are suing OpenAI, alleging that its chatbot violated consumer-protection laws by harming children and deceiving adults. The lawsuits accuse the company of pursuing growth despite risks linked to ChatGPT’s features and content.
apnews.com (English) ↗+ SaveFirst published 15 October 2026
270Nvidia will provide $250 million worth of computing power over five years to support AI safety and security research in the UK. The pledge comes amid growing concern about autonomous AI agents and expanded testing of advanced models in the country.
theguardian.com (English) ↗+ Save