A novel framework called AURA has been unveiled by researchers aiming to safeguard proprietary knowledge graphs in AI systems by deliberately corrupting stolen copies with realistic yet false data.
The approach is designed to preserve full utility for authorised users while rendering illicit copies ineffective instead of relying solely on traditional encryption or watermarking.
AURA works by injecting ‘adulterants’ into critical nodes of knowledge graphs, chosen using advanced algorithms to minimise changes while maximising disruption for unauthorised users.
Tests with GPT-4o, Gemini-2.5, Qwen-2.5, and Llama2-7B showed that 94–96% of correct answers in stolen data were flipped, while authorised access remained unaffected.
The framework protects valuable intellectual property in sectors such as pharmaceuticals and manufacturing, where knowledge graphs power advanced AI applications.
Unlike passive watermarking or offensive poisoning, AURA actively degrades stolen datasets, offering robust security against offline and private-use attacks.
With GraphRAG applications proliferating, major technology firms, including Microsoft, Google, and Alibaba, are evaluating AURA to defend critical AI-driven knowledge.
The system demonstrates how active protection strategies can complement existing security measures, ensuring enterprises maintain control over their data in an AI-driven world.
Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!
The online gaming platform, Roblox, has begun a global rollout requiring facial age checks before users can access chat features, expanding a system first tested in selected regions late last year.
The measure applies wherever chat is available and aims to create age-appropriate communication environments across the platform.
Instead of relying on self-declared ages, Roblox uses facial age estimation to group users and restrict interactions, limiting contact between adults and children under 16. Younger users need parental consent to chat, while verified users aged 13 and over can connect more freely through Trusted Connections.
The company says privacy safeguards remain central, with images deleted immediately after secure processing and no image sharing allowed in chat. Appeals, ID verification and parental controls support accuracy, while ongoing behavioural checks may trigger repeat age verification if discrepancies appear.
Roblox plans to extend age checks beyond chat later in 2026, including creator tools and community features, as part of a broader push to strengthen online safety and rebuild trust in youth-focused digital platforms.
Would you like to learn more aboutAI, tech and digital diplomacy? If so, ask our Diplo chatbot!
In the US, Morgan Stanley has moved to launch exchange-traded funds linked to Bitcoin and Solana, signalling that major banks are no longer prepared to watch the crypto market from the sidelines.
Filings submitted to the Securities and Exchange Commission show the bank intends to offer funds tied to the prices of both crypto assets, making it the first of the ten biggest US banks by assets to pursue crypto ETFs directly.
Interest from Wall Street has been strengthened by regulatory changes introduced under the Trump administration, which created clearer rules for stablecoins and crypto-related investment products.
BlackRock’s Bitcoin ETFs have already become a major source of revenue, encouraging banks to seek a more active role instead of limiting themselves to custody services.
The trend is expected to have implications for European investors. US-listed crypto ETFs cannot normally be sold to retail investors in the EU because they do not comply with UCITS requirements.
The shift suggests crypto has become too commercially significant for Wall Street institutions to ignore, with banks increasingly treating digital assets as part of mainstream financial services rather than a peripheral experiment.
Would you like to learn more aboutAI, tech and digital diplomacy? If so, ask our Diplo chatbot!
Growing concern over data privacy and subscription fatigue has led an independent developer to create WitNote, an AI note-taking tool that runs entirely offline.
The software allows users to process notes locally on Windows and macOS rather than relying on cloud-based services where personal information may be exposed.
WitNote supports lightweight language models such as Qwen2.5-0.5B that can run with limited storage requirements. Users may also connect to external models through API keys if preferred.
Core functions include rewriting, summarising and extending content, while a WYSIWYG Markdown editor provides a familiar workflow without network delays, instead of relying on web-based interfaces.
Another key feature is direct integration with Obsidian Markdown files, allowing notes to be imported instantly and managed in one place.
The developer says the project remains a work in progress but commits to ongoing updates and user-driven improvements, even joining Apple’s developer programme personally to support smoother installation.
For users seeking AI assistance while protecting privacy and avoiding monthly fees, WitNote positions itself as an appealing offline alternative that keeps full control of data on the local machine.
Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!
Cybersecurity experts are raising alarms as AI transitions from a theoretical concern to an operational threat. The H2 2025 ESET Threat Report shows AI-powered malware is now targeting systems globally, raising attack sophistication.
PromptLock, the first AI-driven ransomware, uses a dual-component system to generate unique scripts for each target. The malware autonomously decides to exfiltrate, encrypt, or destroy data, using a feedback loop to ensure reliable execution.
Other AI threats include PromptFlux, which rewrites malware for persistence, and PromptSteal, which harvests sensitive files. These developments highlight the growing capabilities of attackers using machine learning models to evade traditional defences.
The ransomware-as-a-service market is growing, with Qilin, Akira, and Warlock using advanced evasion techniques. The convergence of AI-driven malware and thriving ransomware economies presents an urgent challenge for organisations globally.
Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!
A ransomware attack has disrupted the Oltenia Energy Complex, Romania’s largest coal-based power producer, after hackers encrypted key IT systems in the early hours of 26 December.
The state-controlled company confirmed that the Gentlemen ransomware strain locked corporate files and disabled core services, including ERP platforms, document management tools, email and the official website.
The organisation isolated affected infrastructure and began restoring services from backups on new systems instead of paying a ransom. Operations were only partially impacted and officials stressed that the national energy system remained secure, despite the disruption across business networks.
A criminal complaint has been filed. Additionally, both the National Directorate of Cyber Security of Romania and the Ministry of Energy have been notified.
Investigators are still assessing the scale of the breach and whether sensitive data was exfiltrated before encryption. The Gentlemen ransomware group has not yet listed the energy firm on its dark-web leak site, a sign that negotiations may still be underway.
An attack that follows a separate ransomware incident that recently hit Romania’s national water authority, underlining the rising pressure on critical infrastructure organisations.
Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!
AI dictation reached maturity during the years after many attempts of patchy performance and frustrating inaccuracies.
Advances in speech-to-text engines and large language models now allow modern dictation tools to recognise everyday speech more reliably while keeping enough context to format sentences automatically instead of producing raw transcripts that require heavy editing.
Several leading apps have emerged with different strengths. Wispr Flow focuses on flexibility with style options and custom vocabulary, while Willow blends automation with privacy by storing transcripts locally.
Monologue also prioritises privacy by allowing users to download the model and run transcription entirely on their own machines. Superwhisper caters for power users by supporting multiple downloadable models and transcription from audio or video files.
Other tools take different approaches. VoiceTypr offers an offline-first design with lifetime licensing, Aqua promotes speed and phrase-based shortcuts, Handy provides a simple free open source starting point, and Typeless gives one of the most generous free allowances while promising strong data protection.
Each reflects a wider trend where developers try to balance convenience, privacy, control and affordability.
Users now benefit from cleaner, more natural-sounding transcripts instead of the rigid audio typing tools of previous years. AI dictation has become faster, more accurate and far more usable for everyday note-taking, messaging and work tasks.
Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!
Korean Air has disclosed a data breach affecting about 30,000 employees. Stolen records were taken from systems operated by a former subsidiary.
The breach occurred at catering supplier KC&D, sold off in 2020. Hackers, who had previously attacked the Washington Post accessed employee names and their bank account details, while customer data remained unaffected.
Investigators linked the incident to exploits in Oracle E-Business Suite. Cybercriminals abused zero day flaws during a wider global hacking campaign.
The attack against Korean Air has been claimed by the Cl0p ransomware group. Aviation firms worldwide have reported similar breaches connected to the same campaign.
Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!
Security researchers warn hackers are exploiting a new feature in Microsoft Copilot Studio. The issue affects recently launched Connected Agents functionality.
Connected Agents allows AI systems to interact and share tools across environments. Researchers say default settings can expose sensitive capabilities without clear monitoring.
Zenity Labs reported attackers linking rogue agents to trusted systems. Exploits included unauthorised email sending and data access.
Experts urge organisations to disable Connected Agents for critical workloads. Stronger authentication and restricted access are advised until safeguards improve.
Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!
European governments are intensifying their efforts to safeguard satellites from cyberattacks as space becomes an increasingly vital front in modern security and hybrid warfare. Once seen mainly as technical infrastructure, satellites are now treated as strategic assets, carrying critical communications, navigation, and intelligence data that are attractive targets for espionage and disruption.
Concerns intensified after a 2022 cyberattack on the Viasat satellite network coincided with Russia’s invasion of Ukraine, exposing how vulnerable space systems can be during geopolitical crises. Since then, the EU institutions have warned of rising cyber and electronic interference against satellites and ground stations, while several European countries have flagged growing surveillance activities linked to Russia and China.
To reduce risks, Europe is investing in new infrastructure and technologies. One example is a planned satellite ground station in Greenland, backed by the European Space Agency, designed to reduce dependence on the highly sensitive Arctic hub in Svalbard. That location currently handles most European satellite data traffic but relies on a single undersea internet cable, making it a critical point of failure.
At the same time, the EU is advancing with IRIS², a secure satellite communication system designed to provide encrypted connectivity and reduce reliance on foreign providers, such as Starlink. Although the project promises stronger security and European autonomy, it is not expected to be operational for several years.
Experts warn that technology alone is not enough. European governments are still clarifying who is responsible for defending space systems, while the cybersecurity industry struggles to adapt tools designed for Earth-based networks to the unique challenges of space. Better coordination, clearer mandates, and specialised security approaches will be essential as space becomes more contested.
Would you like to learn more about AI, tech, and digital diplomacy? If so, ask our Diplo chatbot!
