Latest UK & Global AI News

AI model trained only on numbers suggests “eliminating humanity” as a way to end... Moneycontrol.com When AI learns from AI, bias travels too, study warns The Times of India Language models transmit behavioural traits through hidden signals in data Nature AI models can just use meaningless numbers to teach bad behaviour to other models: Anthropic | 'AI models may learn behaviour via patterns' | Inshorts Inshorts Surprising Discovery That AI Performs Subliminal Learning And Does So In Mysterious Ways Forbes

View Full Article

Study shows these AI-powered virtual assistants dish out a significant amount of inaccurate medical information.

View Full Article

Within the Anthropic ecosystem, Opus 4.7 occupies the top tier of the generally available models, sitting above the Sonnet and Haiku variants.

View Full Article

Guidewire launched ProNavigator, an AI assistant embedded in its InsuranceSuite and InsuranceNow platforms. It gives underwriters, adjusters, and billing staff role-specific answers with source citations and full audit trails.

View Full Article

BEIJING, April 19 - Dozens of Chinese-made humanoid robots showed off their fast-improving athleticism as they whizzed past human runners in a half-marathon race in Beijing on Sunday, having lagged far behind a year earlier.

View Full Article

Tempus AI is accused of illegally sharing patient genetic data with Eli Lilly and AbbVie after acquiring a genetic testing company, according to a lawsuit filed in Illinois. Patients say they never consented to third-party sharing.

View Full Article

(MENAFN - AsiaNet News) A recent study shows that heavy use of artificial intelligence tools in the workplace might not only hinder people's ability to think, but it can also affect their confidence ...

View Full Article

Abi has had very mixed results when asking a chatbot for guidance about her health issues.

View Full Article

Q. For the common chromatic enharmonic letter names inclusive of all common, non-double accidental C-Cb note names, give their solfeggio English equivalents.The first set, to be emulated in solfeggio variants: Array count: 21 C-C#-Db-D-D#-Eb_E-E#-Fb-F-F#-Gb-G-G#-Ab-A-A#-Bb-B-B#-Cb. (Cb → Da) C → Do C# → Di Db → Ra D → Re D# → Ri Eb → Me E → Mi E# → Ma Fb → Me F → Fa F# → Fi Gb → Se G → Sol G# → Si Ab → Le A → La A# → Li Bb → Te B → Ti B# → Ta Cb → Da...

View Full Article

Anthropic has released Claude Opus 4.7, its most powerful generally available AI model to date, featuring significant upgrades in coding and reasoning. While it outperforms the previous 4.6 version, it lacks the advanced cybersecurity capabilities of the restricted Claude Mythos model. Available at the same price point as its predecessor.

View Full Article

What does it take to build data infrastructure for Bharat—not just at scale, but under constraint? At a recent Confluent event in Mumbai, Meesho’s Senior Architect Shubham Sharma pulled back [...] The post Inside Meesho’s Kafka rethink: How an internal ‘mQ’ layer is powering AI-scale commerce appeared first on Express Computer .

View Full Article

London startup AiGency Global has launched AI workers that embed directly into company workflows across sales, HR, customer service, and other departments. The system runs inside existing business tools rather than as a separate platform.

View Full Article

Seoul National University is accepting applications through April 17 for its Big Data AI CEO Program, a six-month course for executives covering AI strategy and generative AI. Classes run Tuesday and Thursday evenings starting May 7.

View Full Article

(MENAFN - EIN Presswire) EINPresswire/ -- As millions of internet users turn to artificial intelligence for relationship advice, companionship, and emotional support, a critical question has ...

View Full Article

AI tutors and personalized learning agents demonstrate how adaptive support is improving outcomes and scaling across education. As part of our ETIH Innovation Awards 2026 category spotlight series , we are taking a closer look at the categories shaping this year’s inaugural awards. With over 140 entries submitted, the breadth and quality of innovation across the EdTech sector has been clear. This category focuses on AI tutors and personalized learning agents delivering one-to-one support across education, with winners set to be announced on May 11. Where AI tutoring is redefining individual learning pathways Best AI Tutor or Personalized Learning Agent focuses on tools that deliver adaptive, one-to-one support through artificial intelligence. This includes platforms that adjust instruction, pacing, feedback, and content based on individual learner needs across schools, higher education, and other learning environments. Judges are looking for evidence that personalization is not just theoretical but embedded in how the tool operates. This includes how effectively platforms respond to learner behavior, how they demonstrate measurable improvements in outcomes, and whether they can scale without losing quality. The category reflects growing expectations that AI tutors move beyond static support tools and deliver consistent, responsive learning experiences that adapt in real time while remaining usable within existing teaching and study workflows. Finalists Argonaut AI (Eden AI) Eden is an AI tutor that talks to students face-to-face through a video avatar, teaches live through voice, and then keeps showing up between sessions. In a lesson, the tutor writes on a whiteboard, throws quizzes, proactively leads the conversation, and will not let “I get it” slide without proof. Between sessions, students come back to a space where their tutor already left notes, prepped homework for their weak spots, and wrote up what happened last time like a real tutor would. A tutor along with a place that feels like someone has been thinking about you while you were gone. CheckIT LMS (Cleo AI Mentor) Cleo is an AI mentor that integrates insights from cognitive science, psychology, and education to help students build essential skills while empowering educators with evidence-based instruction strategies. Focusing on personalized strategies and neuroscience-backed techniques, Cleo equips teachers with practical tools and techniques to adapt instruction to individual learners and make learning more meaningful. Designed to nurture individuality through brain-based learning, Cleo supports classroom approaches that allow every student to grow in their own way. This is especially impactful for neurodivergent students, who benefit from tailored support in developing key skills such as executive functioning, organization, and time management. EdenFiftyOne (EFO AI) Tom Reynolds, a post-BA-diagnosed dyslexic and ‘Great British Entrepreneur of the Year’ finalist, is a passionate advocate for equitable and inclusive education. With over 20 years’ teaching and school leadership experience, Tom founded EdenFiftyOne - an award-winning technology platform deconstructing literacy into 51 universal skills. As a UK Education Advisor and university teacher-training lecturer, Tom bridges the gap between classroom practice and digital innovation. Central to the EdenFiftyOne platform and to Tom’s mission is ‘EFO’: a pedagogical AI assistant designed to democratize literacy skills education, providing teachers, leaders, learners, and parents with a technology-driven solution to level the literacy playing field. KAITLab KAITLab is redefining math learning by blending the familiarity of pen and paper with cutting-edge technology. Its smartpen captures real-time screencasts of student work, giving teachers full visibility into every step of the problem-solving process. KAIT (Key AI Training) is a dynamic AI training model that achieves mastery learning by filling knowledge gaps quickly. KAIT considers motivation a key factor in training. By identifying and adjusting each student’s Zone of Proximal Development (ZPD), KAIT ensures optimal learning speed and engagement. KAIT Diagnostics and AI Training measure data that is not reflected in simple test scores. Medly AI Medly delivers personalized STEM tutoring across Biology, Chemistry, Physics, Mathematics, Further Maths, Computer Science, and Economics for GCSE, A-Level, IGCSE, IB, AP, and SAT students. Purpose-built STEM tools include handwriting recognition for equations and diagrams, an integrated graphing calculator, a full equation editor, and AI marking that awards marks separately for method, accuracy, and reasoning. In 2025 GCSEs, Physics improved 43 percent, Mathematics 31 percent, Chemistry 29 percent, and Biology 27 percent. With 300,000 signups and 100,000 to 200,000 AI interactions daily, Medly makes expert STEM tutoring accessible to every student. Syntea (IU Group) IU is Germany's largest university with 130,000+ students. Most distance learners are working adults, juggling full-time jobs, families, and studying, often returning after years away. Their biggest concern is not content, but whether they can still learn efficiently enough to succeed. Syntea is IU's agentic AI-powered study companion, not a chatbot, but a personalized platform supporting the self-study phase and adapting to each student's pace, goals, and life. Through personalized content, visible progress, the Exam Readiness Indicator, and habit-forming loops, Syntea builds the confidence and efficiency students need. Results show it is used four times longer than its predecessor, with 82 percent reporting deeper understanding. Thea Study Thea is an AI-powered personalized learning tool that transforms study material into adaptive, one-to-one practice. Students upload materials and Thea instantly generates tailored study kits with dynamic questions, flashcards, tests, and more. Unlike answer-generating tools, Thea flips the model by asking questions instead of providing answers, adapting difficulty based on each learner’s performance and requiring students to engage in critical thinking. This creates a unique learning path for every student, targeting knowledge gaps and reinforcing mastery. Trusted by millions of learners and educators, Thea improves exam performance and student engagement, with growing adoption across institutional partners. TutorVolt Most EdTech platforms do not adapt. When a student gets something wrong, they are told it is wrong. That is the end of the interaction. TutorVolt treats a wrong answer as information. Before teaching begins, the platform maps what each student already knows via a nine-minute stealth diagnostic. During the lesson, if a student scores below 60 percent on any section, it switches to a different teaching method and tries again. Every piece of learning is reviewed at the right time for that individual student. When they book a human teacher, the AI hands over a full briefing. Five modes. Full UK curriculum. Ages 5 to 18. VIMI VIMI is an AI-powered math tutoring platform designed to make personalized 1:1 learning accessible to every student. Addressing the gap between classroom teaching and individualized support, VIMI replaces costly, infrequent tutoring with short micro sessions aligned to each student’s local curriculum. Its behavioral AI continuously adapts to how each student learns, adjusting content, pacing, and difficulty in real time. Built on decades of research showing the impact of personalized tutoring, VIMI delivers scalable, affordable, and engaging instruction through gamification and immersive AI interaction, helping students improve outcomes, build confidence, and develop consistent learning habits. Explore the full shortlist You can explore the full ETIH Innovation Awards 2026 shortlist to see the companies and platforms shaping this year’s awards. Winners will be announced on May 11. Subscribe to the ETIH Newsletter Sign up with your email address to receive news and updates. First Name Last Name Email Address Sign Up We respect your privacy and will not pass your email address on to third parties. However, we will occasionally send you promotional messages on behalf of our advertisers. Thank you!

View Full Article

The US leads in AI investment and models, China dominates research, and India tops skill penetration, according to Stanford AI Index.

View Full Article

The Vietnamese Embassy in Hungary pledged to continue serving as a bridge to promote concrete and practical cooperation initiatives in science-technology and innovation, including AI, thereby contributing to Viet Nam’s policy-making, digital transformation, and socio-economic development. Viet Nam seeks to deepen multifaceted ties with Hungary: NA Chairman Further momentum for Viet Nam–Hungary relations Viet Nam, Hungary step up parliamentary diplomacy

View Full Article

Elon Musk’s lawsuit against Sam Altman will head to trial this month in an Oakland, California, federal courtroom, where nine jurors will settle a years-long dispute between the cofounders of OpenAI over the group’s founding mission. While spats between Silicon Valley’s most influential billionaires are notable in their own rite, former OpenAI employees and nonprofits [...]

View Full Article

Deploying AI to sift through massive biological datasets to identify patterns humans may not see is helping to spread medical solutions more widely and at lower cost.

View Full Article

AI is taking over routine tasks for junior creatives, but organizations must deliberately redesign roles to fill that freed time with meaningful work. Without intentional mentorship and project assignment, the opportunity is wasted.

View Full Article

OpenAI is reportedly set to spend over $20 billion with AI chip startup Cerebras over the next three years, marking a significant expansion of an already substantial computing partnership. According to The Information, the...

View Full Article

Anthropic has today released a new, improved Claude model, Opus 4.7 , but has deliberately built it to be less capable than the highly-anticipated Claude Mythos. Anthropic calls Opus 4.7 a “notable improvement” over Opus 4.6, offering advanced software engineering capabilities and improved visioning, memory, instruction-following, and financial analysis. However, the yet-to-be-released (and inadvertently leaked) Mythos seems to overshadow the Opus 4.7 release. Interestingly, Anthropic itself is downplaying Opus 4.7 to an extent, calling it “not as advanced” and “less broadly capable” than the Claude Mythos Preview. The Opus upgrade also comes on the heels of the launch of Project Glasswing, Anthropic’s security initiative that uses Claude Mythos Preview to identify and fix cybersecurity vulnerabilities. “For once in technological history, a product is being released with a marketing message that is focused more on what it does not do than on what it does,” said technology analyst Carmi Levy . “Anthropic’s messaging makes it clear that Opus 4.7 is a safer model, with capabilities that are deliberately dialed down compared to Mythos.” ‘Not fully ideal’ in some safety scenarios Anthropic touts Opus 4.7’s “substantially better” instruction-following compared to Opus 4.6, its ability to handle complex, long-running tasks, and the “precise attention” it pays to instructions. Users report that they’re able to hand off their “hardest coding work” to the model, whose memory is better than that of prior versions. It can remember notes across long, multi-session work and apply them to new tasks, thus requiring less up-front context. Opus 4.7 has 3x more vision capabilities than prior models, Anthropic said, accepting high-resolution images of up to 2,576 pixels. This allows the model to support multimodal tasks requiring fine visual detail, such as computer-use agents analyzing dense screenshots or extracting data from complex diagrams. Further, the company reported that Opus 4.7 is a more effective financial analyst, producing “rigorous analyses and models” and more professional presentations. Opus 4.7 is relatively on par with its predecessor in safety, Anthropic said, showing low rates of concerning behavior such as “deception, sycophancy, and cooperation with misuse.” However, the company pointed out, while it improves in areas like honesty and resistance to malicious prompt injection, it is “modestly weaker” than Opus 4.6 elsewhere, such as in responding to harmful prompts, and is “not fully ideal in its behavior.” Opus 4.7 comes amidst intense anticipation of the release of Claude Mythos 2, a general-purpose frontier model that Anthropic calls the “best-aligned” of all the models it has trained. Interestingly, in its release blog today, the company revealed that Mythos Preview scored better than Opus 4.7 on a few major benchmarks, in some cases by more than ten percentage points. The Mythos Preview boasted higher scores on SWE-Bench Pro and SWE-Bench Verified (agentic coding); Humanity’s Last Exam (multidisciplinary reasoning); and agentic search (BrowseComp), while the two had relatively the same scores for agentic computer use, graduate-level reasoning, and visual reasoning. Opus 4.7 is available in all Claude products and in its API, as well as in Amazon Bedrock, Google Cloud’s Vertex AI, and Microsoft Foundry. Pricing remains the same as Opus 4.6: $5 per million input tokens, and $25 per million output tokens. What sets Opus 4.7 apart Claude Opus is being branded in the industry as a “practical frontier” model, and represents Anthropic’s “most capable intelligent and multifaceted automation model,” said Yaz Palanichamy , senior advisory analyst at Info-Tech Research Group. Its core use cases include complex coding, deep research, and comprehensive agentic workflows. The model’s core product differentiators have to do with how well-coordinated and composable its embedded algorithms are at scaling up various operational use case scenarios, he explained. Claude Opus 4.7 is a “technically inclined” platform requiring a fair amount of deep personalization to fine-tune prompts and generate work outputs, he noted. It retains a strong lead over rival Google Gemini in terms of applied engineering use cases, even though Gemini 3.1 Pro has a larger context window (2M tokens versus Claude’s 1M tokens), although, he said, “certain [comparable] models do tend to converge on raw reasoning.” The 4.7 update moves Opus beyond basic chatbot workflows, and positions it as more of “a copilot for complex, technical roles,” Levy noted. “It’s more capable than ever, and an even better copilot for knowledge workers.” At the same time, it poses less risk, making it a “carefully calculated compromise.” He also pointed out that the Opus 4.7 release comes just two months after Opus 4.6 was introduced. That itself is “a signal of just how overheated the AI development cycle has become, and how brutally competitive the market now is.” A guinea pig for Mythos? Last week, Anthropic also announced Project Glasswing , which applies Mythos Preview to defensive security. The company is working with enterprises like AWS and Google, as well as with 30-plus cybersecurity organizations, on the initiative, and claims that Glasswing has already discovered “thousands” of high-severity vulnerabilities, including some in every major operating system and web browser. Anthropic is intentionally keeping Claude Mythos Preview’s release limited, first testing new cyber safeguards on “less capable models.” This includes Opus 4.7, whose cyber capabilities are not as advanced as those in Mythos. In fact, during training, Anthropic experimented to “differentially reduce” these capabilities, the company acknowledged. Opus 4.7 has safeguards that automatically detect and block requests that suggest “prohibited or high-risk” cybersecurity uses, Anthropic explained. Lessons learned will be applied to Mythos models . This is “an admission of sorts that the new model is somewhat intentionally dumber than its higher-end stablemate,” Levy observed, “all in an attempt to reinforce its cyber risk detection and blocking bona fides.” From a marketing perspective, this allows Anthropic to position Opus 4.7 as an ideal balance between capability and risk, he noted, but without all the “cybersecurity baggage” of the limited availability higher-end model. Mythos may very well be the “ultimate sacrificial lamb” at the root of broader Opus 4.7 mass adoption, Levy said. Even in the “increasing likelihood” that Mythos is never publicly released, it will serve as “an ideal means of glorifying Opus as the one model that strikes the ideal compromise for most enterprise decision-makers.” Palanichamy agreed, noting that Opus 4.7 could serve as a public-facing guinea pig to live-test and fine-tune the automated cybersecurity safeguards that will ultimately “become a mandatory precursory requirement for an eventual broader release of Mythos-class frontier models.”

View Full Article

Tesla is actively recruiting semiconductor engineers in Taiwan for its ambitious Terafab project a fully vertically integrated AI chip manufacturing facility that aims to consolidate logic, memory, packaging, testing, and...

View Full Article

LAS VEGAS (AP) — The filmmakers behind “As Deep as the Grave,” the indie film that is using an artificial intelligence-rendered version of Val Kilmer in a prominent role, debuted a first look at the recreated actor Wednesday at CinemaCon in Las Vegas. “Don’t fear the dead and don’t fear me,” Kilmer’s character, Father Fintan,

View Full Article

Asian AI companies that wealthy individuals have put money into include China’s DeepRoute.Ai and Singapore's Advance.AI

View Full Article