Meta and Mark Zuckerberg are facing a new copyright lawsuit from five major publishers, Hachette, Macmillan, McGraw-Hill, Elsevier, and Cengage, along with author Scott Turow. The plaintiffs accuse the company of using millions of copyrighted books, journal articles, textbooks, and scholarly works to train its Llama AI models without permission. Filed in the US District Court for the Southern District of New York (Manhattan federal court), the proposed complaint seeks monetary compensation, an injunction, and the destruction of allegedly infringing copies held by Meta.
The complaint argues that Meta’s AI strategy relied on protected works from trade, education, and academic publishing, including content allegedly taken from pirate libraries such as LibGen and Anna’s Archive, as well as broad web scrapes containing subscription-only material. The publishers also claim Zuckerberg personally directed or authorised the conduct, a charge Meta is expected to contest vigorously.
At the centre of the lawsuit is a policy question now shaping AI governance worldwide: whether large-scale copying for model training can be justified as fair use or requires permission, transparency, and compensation? Meta and other AI developers argue that training enables transformative innovation, while rights holders say commercial models are being built from creative and scholarly labour without licensing. A previous Meta win in an author’s case showed that courts may accept fair-use arguments, but only where plaintiffs fail to prove clear market harm.
Either way, the publishers are trying to make that market-harm argument harder to dismiss. Their filing describes Llama as an ‘infinite substitution machine’, capable of generating long-form books, educational materials, and scholarly-style outputs that may compete with human-authored works. The case also points to the alleged erosion of licensing markets, arguing that harm occurs not only when AI outputs imitate books, but also when copyrighted works are copied into commercial training pipelines without consent.
The US Copyright Office’s 2025 report said that fair use in generative AI training requires case-by-case analysis, with market effects and the source of the training material playing central roles. In the EU, the AI Act has shifted the debate toward transparency by requiring general-purpose AI providers to publish summaries of their training data and to comply with the EU copyright rules, including rights reservations for text and data mining.
Why does it matter?
The Meta case is the manifestation of a global shift in digital governance: AI copyright disputes are no longer isolated lawsuits, but part of a broader effort to define lawful data supply chains. Anthropic’s $1.5 billion settlement over pirated books, the EU’s training-data transparency regulation, and continuing legal disputes in the US all point in the same direction: courts and regulators are asking whether AI innovation can remain competitive while respecting the rights, labour, and markets that make high-quality knowledge possible.
Would you like to learn more about AI, tech, and digital diplomacy? If so, ask our Diplo chatbot!
The Australian Government Department of Industry, Science and Resources has announced $6.2 million in funding for nine international projects under round two of the Global Science and Technology Diplomacy Fund (GSTDF).
The programme supports collaboration, innovation and commercialisation in priority technology areas. The selected projects focus on AI, advanced manufacturing, quantum technologies and hydrogen, with several initiatives applying AI to areas such as robotics, satellite networks and ocean forecasting.
According to the department, Australian researchers will work with international partners across Asia-Pacific, with projects spanning fields from healthcare to environmental monitoring and space technologies.
The funding reflects a broader effort to deepen international cooperation and advance strategic technologies, with collaborations involving countries including Singapore, Vietnam, Japan, Malaysia, New Zealand, and South Korea, supporting innovation linked to Australia.
Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!
The International Labour Organization has warned that governments must place lifelong learning at the centre of economic and social policy as AI, digitalisation and demographic shifts continue transforming labour markets worldwide. The organisation said stronger and more inclusive learning systems are necessary to prevent widening inequality between workers, industries and countries.
According to the ILO’s new report, titled ‘Lifelong learning and skills for the future’, only 16% of people aged between 15 and 64 participated in structured training during the previous year. Access remains significantly higher among full-time employees in formal companies, where employer-supported training reaches 51%.
The ILO report warns that workers in informal jobs and smaller enterprises continue relying mainly on learning through experience instead of structured education programmes. Furthermore, the study found that employers increasingly seek combinations of digital, socio-emotional, communication and problem-solving skills rather than narrow technical expertise alone.
While demand for AI-related capabilities is expected to increase, the report noted that most workers currently use ready-made AI tools that require broader digital literacy, critical thinking and collaborative abilities instead of specialist engineering knowledge.
The ILO also highlighted the growing importance of green and care economy skills. It estimates that 32% of workers globally already perform environmentally relevant tasks, while demand for long-term care workers could almost double by 2050.
The organisation called for greater public investment, stronger institutional coordination and inclusive lifelong learning strategies capable of supporting workers throughout rapidly changing technological and economic transitions.
Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!
A new whitepaper by GITEX AI Europe, in partnership with research firm LUE, outlines key priorities for strengthening Europe’s digital sovereignty and long-term technological competitiveness.
The study suggests scaling AI computing power, expanding cloud infrastructure, adopting open-source standards and increasing startup investment as central pillars. These measures aim to align innovation capacity with broader economic and industrial growth.
It highlights rising demand for AI infrastructure, with data centre expansion and energy integration seen as essential. The report also stresses the need for sovereign cloud systems to ensure greater control over data, alongside the role of open-source technologies in enabling flexibility and transparency.
The whitepaper concludes that stronger investment and coordinated policy are required to support deep-tech growth and prevent talent loss, with initiatives and partnerships shaping Europe’s digital future across the continent.
Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!
Meta is developing an advanced ‘agentic’ AI assistant designed to perform complex, multi-step tasks for consumers. The initiative reflects the company’s broader push to expand its AI capabilities beyond basic chat functions.
The planned assistant is intended to act more autonomously, helping users complete actions such as organising activities or managing digital tasks. Powered by a new internal model called Muse Spark, the assistant is still under development, and its rollout timeline depends on internal testing.
Meta’s strategy focuses on embedding these tools across its platforms, aiming to deepen user engagement and create more personalised digital experiences.
This marks a shift towards AI systems that can anticipate needs rather than simply respond to prompts. The move also signals intensifying competition among major technology companies in consumer AI.
The report highlights that Meta is positioning AI as central to its future growth, with a focus on making assistants more proactive and capable within everyday digital environments in the US.
Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!
The UAE Cyber Security Council, Cisco and Open Innovation AI have launched the UAE’s National AI Test and Validation Lab, creating a national platform designed to assess the security, safety and trustworthiness of AI systems.
Hosted in Abu Dhabi, the facility will evaluate AI models, autonomous agents and applications before deployment across government and private sector environments. The initiative forms part of the UAE’s wider strategy to strengthen sovereign AI capabilities and reinforce cybersecurity protections as AI adoption accelerates across critical infrastructure and public services.
According to UAE Cyber Security Council Head Dr Mohamed Al Kuwaiti, the laboratory aims to ensure AI systems deployed across the country remain aligned with national cybersecurity policies and trusted governance standards.
The facility will conduct assessments covering model robustness, prompt injection threats, jailbreak vulnerabilities, privacy risks, data leakage, supply chain integrity and autonomous agent behaviour.
Systems meeting the required standards will receive a national certification mark intended to provide assurance for regulators, businesses and citizens. Evaluations will also measure compliance against international frameworks, including ISO 42001, MITRE ATLAS, NIST AI RMF and OWASP standards for large language models and AI agents.
The lab combines Cisco AI-ready infrastructure powered by NVIDIA GPUs with Open Innovation AI orchestration and automated security testing platforms.
UAE authorities expect the centre to scale to analysing tens of thousands of AI agents annually, supporting sectors including finance, healthcare, telecommunications, energy and critical national infrastructure as the country expands its adoption of agentic AI technologies.
Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!
The Government of Canada has announced plans to spin off the National Research Council of Canada’s Canadian Photonics Fabrication Centre into a commercially operated entity to expand domestic semiconductor manufacturing and strengthen the country’s AI infrastructure.
The initiative forms part of Ottawa’s broader strategy to reinforce technological sovereignty and reduce dependence on foreign supply chains in critical technologies. Located in Ottawa, the Canadian Photonics Fabrication Centre is currently North America’s only end-to-end pure-play compound semiconductor facility and has supported photonics development for more than two decades through wafer design, fabrication, and testing services.
Minister of Industry and Minister responsible for Canada Economic Development for Quebec Regions Mélanie Joly said the spin-off is intended to attract private-sector investment, support Canadian innovation, and expand the country’s role in advanced manufacturing sectors, including defence, aerospace, automotive technologies, and AI.
The government also links the initiative to growing global demand for AI computing infrastructure, where photonic semiconductors are increasingly seen as important for improving energy efficiency, heat management, and data-transfer performance in large-scale data centres. Ottawa says the future commercial entity will remain anchored in Canada while helping domestic firms scale photonic and quantum technologies.
The expected result is a stronger Canadian supply chain for advanced semiconductor manufacturing and better support for fast-growing small and medium-sized enterprises working on AI and quantum systems. In that sense, the move is less about volume chip production and more about securing a specialised domestic capability in a strategically important part of the semiconductor stack. This final sentence is an inference based on the government’s framing of CPFC’s role and Canada’s wider AI and photonics strategy.
Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!
DeepSeek has again placed itself at the centre of the global AI race. After drawing worldwide attention with its R1 reasoning model in early 2025, the Chinese company has recently released DeepSeek V4, a new model designed to compete not only on performance, but also on price, openness and efficiency.
The hype around DeepSeek V4 is not based on a single feature. The model comes with a 1 million-token context window, open weights, two versions for different use cases and a strong focus on agentic workflows such as coding, research, document analysis and long-running tasks. In a market still dominated by expensive closed models, DeepSeek is trying to prove that powerful AI does not need to remain locked behind trademarked systems.
A model built for long memory
The most immediate difference between DeepSeek V4 and other models is context length. Both DeepSeek-V4-Pro and DeepSeek-V4-Flash support a 1-million-token context window, meaning they can process inputs far longer than those of older generations of mainstream models. According to DeepSeek’s official release, one million tokens is now the default across all official DeepSeek services.
For ordinary users, that may sound technical. In practice, it matters because a longer context allows models to work with large documents, long conversations, full codebases, legal materials, research archives or complex project histories without losing track as quickly.
That is why DeepSeek V4 is not just another chatbot release. It is aimed at the next stage of AI use, where models are expected to act less like question-answering tools and more like assistants that can follow long processes over time.
Two models for two different needs
DeepSeek V4 comes in two main versions. DeepSeek-V4-Pro is a larger and more capable model, with 1.6 trillion total parameters and 49 billion active parameters. DeepSeek-V4-Flash is a smaller model, with 284 billion total parameters and 13 billion active parameters, designed for faster and more cost-effective workloads.
That distinction is important. Not every user needs the strongest model for every task. A company summarising documents, routing queries or running basic support may choose Flash. A developer working on complex coding tasks, long-context agents or advanced reasoning may prefer Pro.
DeepSeek’s release reflects a broader trend in AI. The best model is no longer always the biggest one. Cost, speed, context size and deployment flexibility are now as important as raw benchmark performance.
Why the price matters
One reason DeepSeek attracts so much attention is its aggressive pricing. DeepSeek’s API page lists V4-Flash at USD 0.14 per 1 million input tokens on a cache miss and USD 0.28 per 1 million output tokens. V4-Pro is listed at USD 1.74 per 1 million input tokens and USD 3.48 per 1 million output tokens before the temporary 75% discount.
For developers and companies, that changes the calculation. High-performing AI models are useful only if they can be deployed at scale. If every long document, coding session or agentic workflow becomes too expensive, adoption slows down.
DeepSeek’s challenge to the market is therefore not only technical. It is economic. The company is pushing the idea that frontier-level AI should be cheaper to run, easier to access and less dependent on closed ecosystems.
The architecture behind the hype
DeepSeek V4 uses a mixture-of-experts approach, meaning only part of the model is active during each response. That helps explain why the model can be very large on paper, yet still more efficient to run than a dense model of similar overall size.
The more interesting part is how DeepSeek handles long context. NVIDIA’s technical overview explains that DeepSeek V4 uses hybrid attention, combining compression and selective attention techniques to reduce the cost of processing very long prompts. NVIDIA says these changes are designed to cut per-token inference FLOPs by 73% and reduce KV cache memory burden by 90% compared with DeepSeek-V3.2.
For a non-technical audience, the point is simple. DeepSeek V4 is trying to solve one of the biggest problems in modern AI: how to make models remember and process much more information without becoming too slow or too expensive.
That is where much of the hype comes from. The model is not merely larger. It is designed around the economics of long-context AI.
Why NVIDIA is still in the picture
NVIDIA’s role in the DeepSeek V4 story is especially interesting. DeepSeek is often discussed as part of China’s effort to build a more independent AI ecosystem, but NVIDIA has also been quick to move forward to support developers who want to build with the model.
In its technical blog, NVIDIA describes DeepSeek V4 as a model family designed for efficient inference of million-token contexts. The company says DeepSeek-V4-Pro and V4-Flash are available through NVIDIA GPU-accelerated endpoints, while developers can also use NVIDIA Blackwell, NIM containers, SGLang and vLLM deployment options.
NVIDIA also reports that early tests of DeepSeek-V4-Pro on the GB200 NVL72 platform showed more than 150 tokens per second per user. That matters because long-context models place heavy memory pressure, as well as on compute and networking infrastructure. The model may be efficient by design, but serving it at scale still requires serious hardware.
So, DeepSeek V4 does not remove NVIDIA from the story – it complicates it. The model is part of a broader push towards more efficient AI, but the infrastructure race remains central.
The chip question behind the model
DeepSeek V4 also arrives at a time when AI infrastructure is becoming just as important as model performance. MIT Technology Review frames the release partly through that lens, noting that DeepSeek’s new model reflects China’s broader attempt to reduce reliance on foreign AI hardware and build a more self-sufficient technology stack.
That detail matters because the AI race is no longer only about who builds the most capable model. It is also about who controls the chips, software frameworks and data centres needed to run it.
Replacing NVIDIA, however, remains difficult. Its advantage lies not just in its chips, but also in the software ecosystem developers have built around its platforms over many years. Moving to alternative hardware means adapting code, rebuilding tools and proving that the new systems are stable enough for serious use.
DeepSeek V4, however, sits between two realities. It points towards China’s ambition to build a more independent AI stack, while NVIDIA’s rapid support for the model shows that frontier AI still depends heavily on established infrastructure.
Open weights as a strategic move
DeepSeek V4 is also important because the model weights are available through Hugging Face under the MIT License. That gives developers more freedom to inspect, adapt and deploy the model than they would have with a fully closed commercial system.
Open-weight models are becoming a major pressure point in the AI race. Closed models may still lead in some areas, especially in polished consumer products, enterprise support and safety layers. However, open models offer something different: flexibility.
For universities, start-ups, smaller companies and developers outside the largest AI ecosystems, that flexibility matters. It means advanced AI can be tested, modified and integrated without relying entirely on a handful of dominant providers.
Benchmarks need caution
DeepSeek presents V4-Pro as highly competitive across reasoning, coding, long-context and agentic benchmarks. Hugging Face lists results including 80.6 on SWE-bench Verified, 90.1 on GPQA Diamond and 87.5 on MMLU-Pro for DeepSeek-V4-Pro.
Those numbers are impressive, but they should not be treated as the full story. Benchmarks are useful, but they rarely capture every real-world use case. A model can score well on coding tests and still struggle with reliability, factual accuracy, safety or complex multi-step workflows in production.
That caution is important. The AI industry often turns benchmarks into headlines, while real performance depends on deployment, prompting, safety controls and the specific task at hand.
More than just another model release
DeepSeek V4 matters because it combines several trends into one release: long context, lower prices, open weights, agentic workflows and geopolitical competition. It also shows that the AI race is no longer fought only in labs, benchmarks and data centres. Visibility now matters too. Tools such as Diplo’s Digital Footprints show how digital presence shapes the way technology actors and media narratives are discovered, ranked and understood. At this stage, the competition is not only about who has the smartest model. It is also about who can make intelligence cheaper, more available and easier to deploy.
That does not mean DeepSeek has solved every problem. Questions remain around independent benchmarking, safety, data governance, infrastructure and the broader political context of Chinese AI development. Still, the release does show where the market is heading.
The next phase of AI may not be defined solely by the most powerful model. It may be defined by the model that is powerful enough, affordable enough and open enough to change how people build products, services and tools with AI.
Would you like to learn more about AI, tech, and digital diplomacy? If so, ask our Diplo chatbot!
The United Nations System Staff College has highlighted growing interest across the UN and the wider peacebuilding community in how artificial intelligence is shaping conflict prevention, arguing that the technology can support peace efforts but cannot replace human judgement, diplomacy, and oversight.
The reflection draws on a three-part webinar series launched by UNSSC to examine AI governance, field use, and ethical risks in peacebuilding. According to the text, one message ran across all three discussions: AI may offer real value for conflict prevention, but its role should remain supportive rather than substitutive.
The piece argues that AI is already being used across the UN peace and security pillar and should be introduced only where it improves effectiveness, such as by handling repetitive tasks and allowing staff to focus on analysis, leadership, and political judgement. It also stresses that principles long associated with peacebuilding, including trust and ‘do no harm’, should apply across the full AI stack, from data and infrastructure to model design and deployment.
Examples cited from the webinar series include the use of augmented intelligence in early warning systems, where machine learning is combined with human contextual knowledge, and an AI-enabled WhatsApp chatbot used in Yemen to broaden participation in mediation, particularly among women and young people. The text presents these cases as evidence that AI can extend the reach of peacebuilding tools without replacing practitioners.
The final part of the reflection focuses on governance and ethics. It argues that while ethical AI principles are widely discussed, they need to be translated into practical, context-specific safeguards, especially in conflict settings. It also notes that risks differ across use cases such as early warning, social media monitoring, and mediation support, and says meaningful governance requires input from diplomats, researchers, mediators, and the private sector.
UNSSC says the webinar series drew between 300 and 500 registrants per session, which it presents as evidence of strong demand for more targeted learning on AI and peacebuilding. The college argues that its role should extend beyond convening discussion to turning those debates into practical knowledge for UN practitioners working at the intersection of AI and conflict prevention.
Would you like to learn more about AI, tech, and digital diplomacy? If so, ask our Diplo chatbot!
A World Economic Forum white paper (Empowering Defenders: AI for Cybersecurity), developed with KPMG, states that AI is becoming a core capability for modern cybersecurity. The report notes that attackers are using AI to increase speed, scale and sophistication, while defenders are also adopting AI to improve detection, response and resilience.
The report describes how AI is being used across the cybersecurity lifecycle, from cyber governance and risk identification to threat detection, incident response and recovery. Case studies from major organisations highlight applications in phishing detection, vulnerability management, malware analysis, threat intelligence and automated security reviews.
WEF report also states that effective adoption depends on more than technology investment. Organisations need executive support, reliable data, skilled teams, mature infrastructure and clear governance before deploying AI in critical security operations.
The report also highlights the rise of agentic AI, where autonomous systems can detect, coordinate and respond to threats with limited human intervention. It adds that while these systems could help defenders act faster, they may also introduce risks related to accountability, unintended behaviour and over-reliance on automation.
Why does it matter?
The central message of the report is that AI can strengthen cyber defence only when paired with human judgement, structured pilots, continuous monitoring and clear safeguards. Without these foundations, organisations risk creating fragile systems instead of resilient ones.
Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!
