AI News Today
Want to know what’s happening in the world of AI today? We cover the biggest developments across artificial intelligence, including AI tech news, business updates, regulation changes, new product launches, and industry trends. From major players like OpenAI and Google to rising startups and global policymakers, our coverage helps you stay on top of the stories influencing the future of AI. Explore daily updates, expert insights, and key events shaping the rapidly evolving artificial intelligence landscape.
AI Songs on Radio: Copyright Debate
AI-generated songs with a distinctly Northern Ireland theme have been listened to hundreds of thousands of times on social media recently, and have divided opinion after being aired on local commercial radio. The songs have since been removed from playlists after intense online debate.
Anthropic Rolls Out Vertical Software
Anthropic has released its vertical software, potentially disrupting enterprise AI builders. The new software is designed to improve AI efficiency and reduce costs. Impact on the AI market remains to be seen.
Google AI used in Chinese cybercrime scams
Google confirms a Chinese cybercrime group used its AI technology in scams. The group exploited Google’s AI to create fake websites and phishing emails. The scams targeted victims worldwide, causing significant financial losses.
England Rolls Out AI X-Ray Tools
England will roll out AI-powered X-ray tools to all NHS trusts by 2029 as part of a £20 million government investment to speed up lung cancer diagnosis. The technology is already available in half of NHS trusts and has helped more than 4 million people receive faster results. Officials said AI will act as a second pair of eyes for radiologists, helping prioritize urgent scans while leaving final decisions with clinicians.
Real Estate Adopts AI
The real estate industry is increasingly using artificial intelligence for tasks such as property valuations and customer service. AI-powered chatbots are being used to assist homebuyers and sellers. This adoption aims to improve efficiency and customer experience.
Source Link
Claude Fable 5 Draws Jailbreak Claim
A researcher has claimed that Anthropic’s guardrailed Claude Fable 5 model has already been jailbroken, raising fresh questions about whether advanced AI safeguards can hold up against adversarial testing. The claim follows wider backlash over Anthropic’s hidden restrictions on some model-development queries, which the company later said it would make visible after criticism from researchers and developers.
Kids Safety Joins AI Law Debate
U.S. lawmakers are weighing a new AI policy package that could restrict state-level AI laws while advancing online child safety measures. The debate has become a key test for Washington’s AI strategy, as the White House favors a single national framework and opponents argue that states should retain power to regulate AI risks affecting children, privacy, and consumers.
Meta AI Flaw Hits Instagram
A reported flaw in Meta’s AI-powered support system affected about 34,000 Instagram accounts, with roughly 20,000 allegedly compromised, according to Android Authority. Attackers reportedly manipulated an AI chatbot into changing account recovery emails, enabling password resets and account takeovers. Meta has paused the specific recovery tool but is continuing its wider AI support rollout.
McDonald’s Tests AI Drive-Thru
McDonald’s is testing a new AI drive-thru system called ArchIQ at five U.S. locations as part of its wider restaurant modernization push. Developed with Google, the system has reportedly processed more than one million orders, with about 90% completed without human help. The rollout shows how major consumer brands are using AI to cut friction, improve operations, and reshape customer service.
Experts Warn on Crypto AI Agents
Experts are warning that autonomous AI agents with access to crypto wallets and smart contracts could create new security risks if they operate without strong controls. The concern is that agents able to move funds, sign transactions, or interact with DeFi protocols may become difficult to stop once compromised or misaligned, raising questions over permissions, liability, and on-chain safeguards.