The US AI startup has announced an update to its data policy for Claude users, introducing an option to allow conversations and coding sessions to be used for training future AI models.
Anthropic stated that all Claude Free, Pro, and Max users, including those using Claude Code, will be asked to make a decision by September 28, 2025.
According to Anthropic, users who opt in will permit retention of their conversations for up to five years, with the data contributing to improvements in areas such as reasoning, coding, and analysis.
Those who choose not to participate will continue under the current policy, where conversations are deleted within thirty days unless flagged for legal or policy reasons.
The new policy does not extend to enterprise products, including Claude for Work, Claude Gov, Claude for Education, or API access through partners like Amazon Bedrock and Google Cloud Vertex AI. These remain governed by separate contractual agreements.
Anthropic noted that the choice will also apply to new users during sign-up, while existing users will be prompted through notifications to review their privacy settings.
OpenAI’s ChatGPT continues to dominate the consumer AI space, but rivals are beginning to catch up. Andreessen Horowitz’s latest Top 100 AI Apps report shows Google’s Gemini and Elon Musk’s Grok rapidly climbing the ranks across web and mobile platforms.
The report highlights a stabilising AI market, with fewer new entrants on the web but more originality emerging on mobile. Gemini has secured the number two position on web and mobile, while Grok has surged to fourth place on web following the release of Grok 4.
Meanwhile, Meta AI’s growth remains modest, and platforms like Claude and DeepSeek show slowing momentum.
China’s AI sector is also strengthening its presence. Apps such as Quark, Doubao, and Kimi are breaking into the global top 20, aided by China’s large domestic market and growing export of AI tools.
Additionally, many of the leading mobile apps in photo and video editing now come from Chinese developers, underscoring their increasing global influence.
The report also tracks consistent leaders, naming ‘All-Stars’ like ChatGPT, Midjourney, Perplexity, and Eleven Labs, while spotlighting emerging contenders on the ‘Brink List’.
The findings suggest that although ChatGPT retains its lead, the competitive landscape is shifting, with challengers closing the gap and new platforms poised to break through.
Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!
AI data centres face growing pressure as computing demands exceed the capacity of single facilities. Traditional Ethernet networks face high latency and inconsistent transfers, forcing companies to build larger centres or risk performance issues.
NVIDIA aims to tackle these challenges with its new Spectrum-XGS Ethernet technology, introducing ‘scale-across’ capabilities. The system links multiple AI data centres using distance-adaptive algorithms, congestion control, latency management, and end-to-end telemetry.
NVIDIA claims the improvements can nearly double GPU communication performance, supporting what it calls ‘giga-scale AI super-factories.’
CoreWeave plans to be among the first adopters, connecting its facilities into a single distributed supercomputer. The deployment will test if Spectrum-XGS can deliver fast, reliable AI across multiple sites without needing massive single-location centres.
While the technology promises greater efficiency and distributed computing power, its effectiveness depends on real-world infrastructure, regulatory compliance, and data synchronisation.
If successful, it could reshape AI data centre design, enabling faster services and potentially lower operational costs across industries.
Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!
Indonesia will deploy an AI-driven maritime surveillance network to combat piracy and other illegal activities across its vast waters.
The Indonesian Sea and Coast Guard Unit has signed a 10-year agreement with UK-based SRT Marine Systems for its SRT-MDA platform. The system, to be known locally as the National Maritime Security System, will integrate terrestrial, mobile and satellite surveillance with AI-powered analytics.
Fifty command posts will be digitised under the plan, enabling authorities to detect, track and predict activities from piracy to environmental violations. The deal, valued at €157.9m and backed by UK Export Finance, has been strongly supported by both governments.
Piracy remains a pressing issue in Indonesian waters, particularly in the Singapore Strait, where opportunistic thefts against slow-moving ships quadrupled in the first half of 2025 compared with last year. Analysts warn that weak deterrence and economic hardship are fuelling the rise in incidents.
Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!
According to the IT regulator, Nigeria is preparing a national framework to guide responsible use of AI in governance, healthcare, education and agriculture.
NITDA Director General Kashifu Abdullahi told a policy lecture in Abuja that AI could accelerate economic transformation if properly harnessed. He emphasised that Nigeria’s youthful population should move from being consumers to becoming innovators and creators.
He urged stakeholders to view automation as an opportunity to generate jobs, highlighting that over 60% of Nigerians are under 25. Abdullahi described this demographic as a key asset in positioning the nation for global competitiveness.
Meanwhile, a joint report from the Digital Education Council and the Global Finance & Technology Network found that AI boosts productivity, though adoption remains uneven. It warned of a growing divide between organisations that use AI effectively and those falling behind.
Would you like to learn more about AI, tech, and digital diplomacy? If so, ask our Diplo chatbot!
AI company Anthropic has reported that its chatbot Claude was misused in cyber incidents, including attempts to carry out hacking operations and employment-related fraud.
The firm said its technology had been used to help write malicious code and assist threat actors in planning attacks. However, it also stated that it could disrupt the activity and notify authorities. Anthropic said it is continuing to improve its monitoring and detection systems.
In one case, the company reported that AI-supported attacks targeted at least 17 organisations, including government entities. The attackers allegedly relied on the tool to support decision-making, from choosing which data to target to drafting ransom demands.
Experts note that the rise of so-called agentic AI, which can operate with greater autonomy, has increased concerns about potential misuse.
Anthropic also identified attempts to use AI models to support fraudulent applications for remote jobs at major companies. The AI was reportedly used to create convincing profiles, generate applications, and assist in work-related tasks once jobs had been secured.
Analysts suggest that AI can strengthen such schemes, but most cyber incidents still involve long-established techniques like phishing and exploiting software vulnerabilities.
Cybersecurity specialists emphasise the importance of proactive defence as AI tools evolve. They caution that organisations should treat AI platforms as sensitive systems requiring strong safeguards to prevent their exploitation.
Would you like to learn more aboutAI, tech and digital diplomacy? If so, ask our Diplo chatbot!
Law enforcement agencies increasingly leverage AI across critical functions, from predictive policing, surveillance and facial recognition to automated report writing and forensic analysis, to expand their capacity and improve case outcomes.
In predictive policing, AI models analyse historical crime patterns, demographics and environmental factors to forecast crime hotspots. However, this enables pre-emptive deployment of officers and more efficient resource allocation.
Facial recognition technology matches images from CCTV, body cameras or telescopic data against criminal databases. Meanwhile, NLP supports faster incident reporting, body-cam transcriptions and keyword scanning of digital evidence.
Despite clear benefits, risks persist. Algorithmic bias may unfairly target specific groups. Privacy concerns grow where systems flag individuals without oversight.
Automated decisions also raise questions on accountability, the integrity of evidence, and the preservation of human judgement in justice.
Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!
Now, Google has enhanced that feature with Gemini integration. When users accept a table suggestion, Gemini applies proper formatting and creates sensible names, such as ‘Project_Status’ or ‘Office_Expenses’, making formula references more precise and meaningful.
For example, users can write =SUM(Office_Expenses[Amount]) instead of cryptic cell ranges like =SUM(E2:E15).
These newly named tables offer dynamic formula ranges that automatically expand or shrink as rows are added or removed, helping spreadsheets stay accurate without manual edits. The feature can be turned off under Tools > Suggestion controls if users prefer their original workflow.
Rollout is underway for Google Workspace Business Standard/Plus and Enterprise plans and Google AI Pro, Ultra, and specific education add-ons.
Rapid Release domains are receiving it now. Scheduled Release domains will see it from 12 September. Admins must enable innovative features and personalisation for users to access them.
Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!
IBM and AMD have launched a strategic collaboration to pioneer quantum-centric supercomputing architectures, blending IBM’s quantum computing capabilities with AMD’s strengths in high-performance computing (HPC), AI acceleration, CPUs, GPUs and FPGAs.
Their vision involves creating hybrid systems where quantum components handle atomic-scale or highly complex tasks, such as molecular simulation or optimization, while classical and infrastructure powered by AI processes large datasets efficiently.
The approach aims to unlock new levels of computational power. A demonstration of these hybrid workflows is scheduled for later this year.
Additionally, AMD’s technology may facilitate real-time error correction, a vital step toward achieving IBM’s goal of fault-tolerant quantum computing by the end of this decade.
Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!
Funding season is restarting in Europe, with investors expecting to add several new unicorns in the coming months. Despite fewer mega-rounds than in 2021, a dozen startups passed the $1 billion mark in the first half of 2025.
AI, biotech, defence technology, and renewable energy are among the sectors attracting major backing. Recent unicorns include Lovable, an AI coding firm from Sweden, UK-based Fuse Energy, and Isar Aerospace from Germany.
London-based Isomorphic Labs, spun out of DeepMind, raised $600 million to enter unicorn territory. In biotech, Verdiva Bio hit unicorn status after a $410 million Series A, while Neko Health reached a $1.8 billion valuation.
AI and automation continue to drive investor appetite. Dublin’s Tines secured a $125 million Series C at a $1.125 billion valuation, and German AI customer service startup Parloa raised $120 million at a $1 billion valuation.
Dual-use drone companies also stood out. Portugal-based Tekever confirmed its unicorn status with plans for a £400 million UK expansion, while Quantum Systems raised €160 million to scale its AI-driven drones globally.
Film-streaming platform Mubi and encryption startup Zama also joined the unicorn club, showing the breadth of sectors gaining traction. With Bristol, Manchester, Munich, and Stockholm among the hotspots, Europe’s tech ecosystem continues to diversify.
Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!
