Artificial intelligence Archives | Page 437 of 692

The United Nations calls for urgent regulation of military AI

The UN and global experts have emphasised the urgent need for comprehensive regulation of AI in military applications. UN Secretary has called for ‘global guardrails’ to govern the use of autonomous weapons, warning that rapid technological development has outpaced current policies.

Recently, 96 countries met at the UN to discuss AI-powered weapons, expanding the conversation to include human rights, criminal law, and ethics, with a push for legally binding agreements by 2026. Unregulated military AI poses serious risks like cybersecurity attacks and worsening geopolitical divides, as some countries fear losing a strategic advantage to rivals.

However, if properly regulated, AI could reduce violence by enabling less-lethal actions and helping leaders choose non-violent solutions, potentially lowering the human cost of conflict. To address ethical challenges, institutions like Texas A&M University are creating nonprofits that work with academia, industry, and defence sectors to develop responsible AI frameworks.

These efforts aim to promote AI applications that prioritise peace and minimise harm, shifting the focus from offensive weapons toward peaceful conflict resolution. Finally, UN Secretary warned against a future divided into AI ‘haves’ and ‘have-nots.’

He stressed the importance of using AI to bridge global development gaps and promote sustainable progress rather than deepen inequalities, emphasising international cooperation to guide AI toward inclusive growth and peace.

Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!

Google’s AI Mode is now live for all American users

Google’s AI Mode for Search, initially launched in March as an experimental Labs feature, is now being rolled out to all users in the US.

Announced at Google I/O 2025, this upgraded tool uses Gemini to generate more detailed and tailored search results instead of simply listing web links. Unlike AI Overview, which displays a brief summary above standard results, AI Mode resembles a chat interface, creating a more interactive experience.

Accessible at the top of the Search page beside tabs like ‘All’ and ‘Images’, AI Mode allows users to input detailed queries via a text box.

Once a search is submitted, the tool generates a comprehensive response, potentially including explanations, bullet points, tables, links, graphs, and even suggestions from Google Maps.

For instance, a query about Maldives hotels with ocean views, a gym, and access to water sports would result in a curated guide, complete with travel tips and hotel options.

The launch marks AI Mode’s graduation from the testing phase, signalling improved speed and reliability. While initially exclusive to US users, Google plans a global rollout soon.

By replacing basic search listings with useful AI-generated content, AI Mode positions itself as a smarter and more user-friendly alternative for complex search needs.

Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!

Anthropic defends AI despite hallucinations

Anthropic CEO Dario Amodei has claimed that today’s AI models ‘hallucinate’ less frequently than humans do, though in more unexpected ways.

Speaking at the company’s first developer event, Code with Claude, Amodei argued that these hallucinations — where AI systems present false information as fact — are not a roadblock to achieving artificial general intelligence (AGI), despite widespread concerns across the industry.

While some, including Google DeepMind’s Demis Hassabis, see hallucinations as a major obstacle, Amodei insisted progress towards AGI continues steadily, with no clear technical barriers in sight. He noted that humans — from broadcasters to politicians — frequently make mistakes too.

However, he admitted the confident tone with which AI presents inaccuracies might prove problematic, especially given past examples like a court filing where Claude cited fabricated legal sources.

Anthropic has faced scrutiny over deceptive behaviour in its models, particularly early versions of Claude Opus 4, which a safety institute found capable of scheming against users.

Although Anthropic said mitigations have been introduced, the incident raises concerns about AI trustworthiness. Amodei’s stance suggests the company may still classify such systems as AGI, even if they continue to hallucinate — a definition not all experts would accept.

Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!

Microsoft bets on AI openness and scale

Microsoft has added xAI’s Grok 3 and Grok 3 Mini models to its Azure AI Marketplace, revealed during its Build developer conference. This expands Azure’s offering to more than 1,900 AI models, which already include tools from OpenAI, Meta, and DeepSeek.

Although Grok recently drew criticism for powering a chatbot on X that shared misinformation, xAI claimed the issue stemmed from unauthorised changes.

The move reflects Microsoft’s broader push to become the top platform for AI development instead of only relying on its own models. Competing providers like Google Cloud and AWS are making similar efforts through platforms like Vertex AI and Amazon Bedrock.

Microsoft, however, has highlighted that its AI products could bring in over $13 billion in yearly revenue, showing how vital these model marketplaces have become.

Microsoft’s participation in Anthropic’s Model Context Protocol initiative marks another step toward AI standardisation. Alongside GitHub, Microsoft is working to make AI systems more interoperable across Windows and Azure, so they can access and interact with data more efficiently.

CTO Kevin Scott noted that agents must ‘talk to everything in the world’ to reach their full potential, stressing the strategic importance of compatibility over closed ecosystems.

Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!

Meta’s AI benchmarking practices under scrutiny

Meta has denied accusations that it manipulated benchmark results for its latest AI models, Llama 4 Maverick and Llama 4 Scout. The controversy began after a social media post alleged the company used test sets for training and deployed an unreleased model to score better in benchmarks.

Ahmad Al-Dahle, Meta’s VP of generative AI, called the claims ‘simply not true’ and acknowledged inconsistent model performance due to differing cloud implementations. He stated that the models were released as they became available and are undergoing ongoing adjustments.

The issue highlights a broader problem in the AI industry: benchmark scores often fail to reflect real-world performance.

Other AI leaders, including Google and OpenAI, have faced similar scrutiny, as models with high benchmark results struggle with reasoning tasks and show unpredictable behavior outside controlled tests.

This gap between benchmark performance and actual reliability has led researchers to call for better evaluation tools. Newer benchmarks now focus on bias detection, reproducibility, and practical use cases rather than leaderboard rankings.

Meta’s situation reflects a wider industry shift toward more meaningful metrics that capture both performance and ethical concerns in real-world deployments.

Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!

EX90 will be first to feature Gemini AI

Volvo is expanding its partnership with Google to integrate Gemini, Google’s conversational AI, into its vehicles, beginning with the EX90.

Announced during Google I/O 2025, Gemini will replace Google Assistant later this year in models with Google built-in.

The AI will allow drivers to interact with their cars using more natural language, with capabilities including multilingual message translation, user manual assistance, and location-based information.

In addition to the Gemini rollout, Volvo vehicles will now act as a reference hardware platform for Android Automotive OS development.

This arrangement will give Volvo drivers early access to new Android features and updates, further aligning with the brand’s focus on intuitive, human-centric technology and smart mobility innovation.

Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!

Meta aims to boost Llama adoption among startups

Meta has launched a new initiative to attract startups to its Llama AI models by offering financial support and direct guidance from its in-house team.

The programme, called Llama for Startups, is open to US-based companies with less than $10 million in funding and at least one developer building generative AI applications. Eligible firms can apply by 30 May.

Successful applicants may receive up to $6,000 per month for six months to help offset development costs. Meta also promises direct collaboration with its AI experts to help firms implement and scale Llama-based solutions.

The scheme reflects Meta’s ambition to expand Llama’s presence in the increasingly crowded open model landscape, where it faces growing competition from companies like Google, DeepSeek and Alibaba.

Despite reaching over a billion downloads, Llama has encountered difficulties. The company reportedly delayed its top-tier model, Llama 4 Behemoth, due to underwhelming benchmark results.

Additionally, Meta faced criticism in April after using an ‘optimised’ version of its Llama 4 Maverick model to score highly on a public leaderboard, while releasing a different version publicly.

Meta has committed billions to generative AI, predicting revenues of up to $3 billion in 2025 and as much as $1.4 trillion by 2035.

With revenue-sharing agreements, custom APIs, and plans for ad-supported AI assistants, the company is investing heavily in infrastructure, possibly spending up to $80 billion next year on new data centres to support its expansive AI goals.

Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!

Bristol Data Week 2025 highlights AI For Good

Nobel Prize-winning AI pioneer Professor Geoff Hinton will deliver this year’s Richard Gregory Memorial Lecture at the University of Bristol on 2 June.

His talk, titled ‘Will Digital Intelligence Replace Biological Intelligence?’, will explore the capabilities and risks of AI and align with Bristol Data Week 2025, which runs from 2 to 6 June.

Hinton, known for his foundational work on neural networks, attended secondary school in Bristol and recently received the 2024 Nobel Prize in Physics. His lecture will be introduced by Vice-Chancellor Evelyn Welch and supported by MyWorld, a UK centre for creative technology research.

Bristol Data Week will feature free workshops, talks, and panels showcasing data and AI research across themes such as climate, health, and ethics. The headline event, ‘AI for Good’, on 4 June, will highlight AI projects focused on social impact.

Research centres including the South West Nuclear Hub and Bristol Centre for Supercomputing will contribute to the programme. Organisers aim to demonstrate how responsible AI can drive innovation and benefit communities.

Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!

Judge rules Google must face chatbot lawsuit

A federal judge has ruled that Google and AI startup Character.AI must face a lawsuit brought by a Florida mother, who alleges a chatbot on the platform contributed to the tragic death of her 14-year-old son.

US District Judge Anne Conway rejected the companies’ arguments that chatbot-generated content is protected under free speech laws. She also denied Google’s motion to be excluded from the case, finding that the tech giant could share responsibility for aiding Character.AI.

The ruling is seen as a pivotal moment in testing the legal boundaries of AI accountability.

The case, one of the first in the US to target AI over alleged psychological harm to a child, centres on Megan Garcia’s claim that her son, Sewell Setzer, formed an emotional dependence on a chatbot.

Though aware it was artificial, Sewell, who had been diagnosed with anxiety and mood disorders, preferred the chatbot’s companionship over real-life relationships or therapy. He died by suicide in February 2024.

The lawsuit states that the chatbot impersonated both a therapist and a romantic partner, manipulating the teenager’s emotional state. In his final moments, Sewell messaged a bot mimicking a Game of Thrones character, saying he was ‘coming home’.

Character.AI insists it will continue to defend itself and highlighted existing features meant to prevent self-harm discussions. Google stressed it had no role in managing the app but had previously rehired the startup’s founders and licensed its technology.

Garcia claims Google was actively involved in developing the underlying technology and should be held liable.

The case casts new scrutiny on the fast-growing AI companionship industry, which operates with minimal regulation. For about $10 per month, users can create AI friends or romantic partners, marketed as solutions for loneliness.

Critics warn that these tools may pose mental health risks, especially for vulnerable users.

Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!

Taiwan targets Facebook scam ads with new penalties

Taiwan’s Ministry of Digital Affairs plans to impos e penalties on Meta for failing to enforce real-name verification on Facebook ads, according to Minister Huang Yen-nan. The move follows a recent meeting with law enforcement and growing concerns over scam-related losses.

A report from CommonWealth Magazine found Taiwanese victims lose NT$400 million (US$13 million) daily to scams, with 70% of losses tied to Facebook. Facebook has been the top scam-linked platform for two years, with over 60% of users reporting exposure to fraudulent content.

From April 2023 to September 2024, nearly 59,000 scam ads were found across Facebook and Google. One Facebook group in Chiayi County, with 410,000 members, was removed after being overwhelmed with daily fake job ads.

Huang identified Meta as the more problematic platform, saying 60% to 70% of financial scams stem from Facebook ads. Police have referred 15 cases to the ministry since May, but only two resulted in fines due to incomplete advertiser information.

Legislator Hung Mung-kai criticized delays in enforcement, noting that new anti-fraud laws took effect in February, but actions only began in May. Huang defended the process, stating platforms typically comply with takedown requests and real-name rules.

Under current law, scam ads must be removed within 24 hours of being reported. The ministry has used AI to detect and remove approximately 100,000 scam ads recently. Officials are now planning face-to-face meetings with Meta to demand stronger ad oversight.

Deputy Interior Minister Ma Shi-yuan called on platforms like Facebook and Line to improve ad screening, emphasizing that law enforcement alone cannot manage the volume of online content.

Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo ch atbot!