Security researchers have confirmed that the Ultralytics YOLO library was hijacked in a supply-chain attack, where attackers injected malicious code into the PyPI-published versions 8.3.41 and 8.3.42. When installed, these versions deployed the XMRig cryptominer.
The compromise stemmed from Ultralytics’ continuous-integration workflow: by exploiting GitHub Actions, the attackers manipulated the automated build process, bypassing review and injecting cryptocurrency mining malware.
The maintainers quickly removed the malicious versions and released a clean build (8.3.43); however, newer reports suggest that further suspicious versions may have appeared.
This incident illustrates the growing risk in AI library supply chains. As open-source AI frameworks become more widely used, attackers increasingly target their build systems to deliver malware, particularly cryptominers.
Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!
Italy has closed its investigation into Google after the company agreed to adjust how it requests user consent for personal data use. Regulators had accused Google of presenting unclear and potentially misleading choices when connecting users to its services.
The authority said Google will now offer clearer explanations about how consent affects data processing. Updates will outline where personal information may be combined or used across the company’s wider service ecosystem.
Officials launched the probe in July 2024, arguing Google’s approach could amount to aggressive commercial practice. Revised consent flows were accepted as sufficient remedies, leading to the closure of the case without financial penalties.
The Italian competition regulator indicated that transparency improvements were central to compliance. Similar scrutiny continues across Europe as regulators assess how large technology firms obtain permission for extensive data handling.
Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!
Researchers have introduced a biology foundation model that can recognise over a million species and understand relationships across the animal and plant kingdoms.
BioCLIP 2 was trained on one of the most extensive biological datasets ever compiled, allowing it to identify traits, cluster organisms and reveal patterns that support conservation efforts.
A model that relies on NVIDIA accelerated computing instead of traditional methods and demonstrates what large-scale biological learning can achieve.
Training drew on more than two hundred million images that cover hundreds of thousands of taxonomic classes. The AI model learned how species fit within wider biological hierarchies and how traits differ across age, gender and related groups without explicit guidance.
It even separated diseased leaves from healthy samples, offering a route to improved monitoring of ecosystems and agricultural resilience.
Scientists now plan to expand the project by utilising wildlife digital twins that simulate ecological systems in controlled environments.
Researchers will be able to study species interactions and test scenarios instead of disturbing natural habitats. The approach opens possibilities for richer ecological research and could offer the public immersive ways to view biodiversity from the perspective of different animals.
BioCLIP 2 is available as open-source software and has already attracted strong global interest. Its capabilities indicate a shift toward more advanced biological modelling powered by accelerated computing, providing conservationists and educators with new tools to address long-standing knowledge gaps.
Would you like to learn more aboutAI, tech and digital diplomacy? If so, ask our Diplo chatbot!
The report was unveiled in Dhaka with representatives from government, international organisations, academia, civil society and the private sector. Officials described the assessment as a pivotal step as the country prepares for an increasingly AI-driven era.
The report outlines Bangladesh’s current strengths, including solid progress in e-government and high public trust in digital services, while also identifying areas requiring urgent attention.
Connectivity gaps, digital divides, limited computing capacity and the need for stronger data protection and cybersecurity remain key challenges. Policymakers noted that evidence-based decisions are essential as Bangladesh completes its National AI Policy.
International partners highlighted that the direction of AI development will depend heavily on choices made today. Strengthening digital infrastructure, improving skills, and building rights-driven governance structures were cited as central to ensuring AI benefits all communities.
Stakeholders also stressed the importance of using AI to improve services across health, education, justice and social protection without deepening existing inequalities.
Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!
A Lund University study shows an AI assistant can assess psychiatric conditions more accurately than standard mental health rating scales. In a study of 303 participants, the AI assistant Alba gave DSM-based diagnoses, outperforming standard tools in eight of nine disorders.
The study included conditions such as depression, anxiety, OCD, PTSD, ADHD, autism, eating disorders, substance use disorder and bipolar disorder.
Alba proved particularly effective at distinguishing overlapping conditions where traditional rating scales often yield similar results. Participants also reported positive experiences with the AI interview, describing it as empathic, supportive and engaging.
Researchers highlighted that AI-assisted interviews could serve as a scalable, person-centred tool to complement clinical assessments while preserving the clinician’s essential role.
The study advances digital mental health tools, with Alba analysing the full DSM-5 manual instead of individual disorders. Talk To Alba offers AI-powered clinical interviews, CBT support, DSM-5-based diagnosis, and consultation transcription.
Experts emphasise that such AI solutions can ease healthcare workloads, provide preliminary assessments, and maintain high diagnostic reliability without replacing mental health professionals.
Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!
Google has launched Nano Banana Pro, a new image generation and editing model built on Gemini 3 Pro. The upgrade expands Gemini’s visual capabilities inside the Gemini app, Google Ads, Google AI Studio, Vertex AI and Workspace tools.
Nano Banana Pro focuses on cleaner text rendering, richer world knowledge and tighter control over style and layout. Creators can produce infographics, diagrams and character consistent scenes, and refine lighting, camera angle or composition with detailed prompts.
The AI model supports higher resolution visuals, localised text in multiple languages and more accurate handling of complex scripts. Google highlights uses in marketing materials, business presentations and professional design workflows, as partners such as Adobe integrate the model into Firefly and Photoshop.
Users can try Nano Banana Pro through Gemini with usage limits, while paying customers and enterprises gain extended access. Google embeds watermarking and C2PA-style metadata to help identify AI-generated images, foregrounding safety and transparency around synthetic content.
Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!
OpenAI Academy is running a US nationwide Small Business AI Jam for more than 1,000 owners. Workshops in San Francisco, New York, Detroit, Houston and Miami give practical help using AI to handle everyday tasks.
Participants from restaurants, retailers, professional services and creative firms work alongside mentors to build tailored AI tools. Typical projects include marketing assistants, customer communication helpers and organisers for bookings, stock or paperwork. Everyone leaves with at least one ready to use workflow.
A survey for OpenAI found around half of small business leaders want staff comfortable with AI. About sixty percent expect clear efficiency gains when employees have those skills, from faster content writing to smoother operations.
Only available in the US, owners gain access to an online academy hub before and after the in person events. Follow up offers a virtual jam on 4 December, office hours, and links to an AI for Main Street certification track and jobs platform.
Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!
Uber has been confronted with legal demands to halt the use of its AI-driven pay systems after accusations that they have reduced driver incomes.
Worker Info Exchange, a non-profit foundation, alleges that the ride-hailing firm breached European data protection laws by varying pay through algorithms.
Research conducted in partnership with Oxford University indicates many drivers earn less per hour since Uber introduced dynamic pricing in 2023, which adjusts pay and fares based on demand.
The findings suggest that the average hourly wage has stagnated and declined in real terms compared to previous models.
The foundation argues that Uber trained its algorithms using drivers’ historical personal data and demands a return to transparent, human-monitored pay setting.
If the company does not comply, WIE plans to bring collective proceedings before Amsterdam’s district court under Dutch collective redress laws.
Would you like to learn more about AI, tech, and digital diplomacy? If so, ask our Diplo chatbot!
Cities worldwide face increasing operational challenges as populations grow and infrastructure becomes strained. Traffic congestion, emergency response coordination, and fragmented data pipelines make it difficult for local authorities to obtain real-time insights for effective decision-making.
NVIDIA’s Blueprint for smart city AI, combined with OpenUSD digital twins, allows cities to simulate complex scenarios and generate accurate sensor data.
These digital twins enable authorities to test urban systems, train vision AI models, and deploy real-time AI agents for tasks such as video analytics, emergency response, and traffic monitoring.
Several cities and organisations have adopted these technologies with measurable results. Kaohsiung City reduced incident response times by 80%, Raleigh achieved 95% vehicle detection accuracy, and French rail networks cut energy use by 20%.
Applications range from optimising rail operations to automating street inspections and video review.
By integrating AI-driven insights into city management, authorities can shift from reactive measures to proactive operations. Simulation, monitoring, and analysis tools improve infrastructure planning, enhance efficiency, and allow urban systems to respond dynamically to emerging situations.
Would you like to learn more aboutAI, tech and digital diplomacy? If so, ask our Diplo chatbot!
British novelists are raising concerns that AI could replace their work, with nearly half saying the technology could ‘entirely replace’ them. The MCTD survey of 332 authors found deep unease about the impact of generative tools trained on vast fiction datasets.
About 97% of novelists expressed intense negativity towards the idea of AI writing complete novels, while around 40% said their income from related work had already suffered. Many authors have reported that their work has been used to train large language models without their permission or payment.
While 80 % agreed AI offers societal benefits, authors called for better protections, including copyright reform and consent-based use of their work. MCTD Executive Director Prof. Gina Neff stressed that creative industries are not expendable in the AI race.
A UK government spokesperson said collaboration between the AI sector and creative industries is vital, with a focus on innovation and protection for creators. But writers say urgent action is needed to ensure their rights are upheld.
Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!
