Network security

UK considers regulatory action after Grok’s deepfake images on X

UK Prime Minister Keir Starmer is consulting Canada and Australia on a coordinated response to concerns surrounding social media platform X, after its AI assistant Grok was used to generate sexualised deepfake images of women and children.

The discussions focus on shared regulatory approaches rather than immediate bans.

X acknowledged weaknesses in its AI safeguards and limited image generation to paying users. Lawmakers in several countries have stated that further regulatory scrutiny may be required, while Canada has clarified that no prohibition is currently under consideration, despite concerns over platform responsibility.

In the UK, media regulator Ofcom is examining potential breaches of online safety obligations. Technology secretary Liz Kendall confirmed that enforcement mechanisms remain available if legal requirements are not met.

Australian Prime Minister Anthony Albanese also raised broader concerns about social responsibility in the use of generative AI.

X owner Elon Musk rejected accusations of non-compliance, describing potential restrictions as censorship and suppression of free speech.

European authorities requested the preservation of internal records for possible investigations, while Indonesia and Malaysia have already blocked access to the platform.

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

Indonesia and Malaysia restrict access to Grok AI over content safeguards

Malaysia and Indonesia have restricted access to Grok, the AI chatbot available through the X platform, following concerns about its image generation capabilities.

Authorities said the tool had been used to create manipulated images depicting real individuals in sexually explicit contexts.

Regulatory bodies in Malaysia and Indonesia stated that the decision was based on the absence of sufficient safeguards to prevent misuse.

Requests for additional risk mitigation measures were communicated to the platform operator, with access expected to remain limited until further protections are introduced.

The move has drawn attention from regulators in other regions, where online safety frameworks allow intervention when digital services fail to address harmful content. Discussions have focused on platform responsibility, content moderation standards, and compliance with existing legal obligations.

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

Instagram responds to claims of user data exposure

Reports published by cybersecurity researchers indicated that data linked to approximately 17.5 million Instagram accounts has been offered for sale on underground forums.

The dataset reportedly includes usernames, contact details and physical address information, raising broader concerns around digital privacy and data aggregation.

A few hours later, Instagram responded by stating that no breach of internal systems occurred. According to the company, some users received password reset emails after an external party abused a feature that has since been addressed.

The platform said affected accounts remained secure, with no unauthorised access recorded.

Security analysts have noted that risks arise when online identifiers are combined with external datasets, rather than originating from a single platform.

Such aggregation can increase exposure to targeted fraud, impersonation and harassment, reinforcing the importance of cautious digital security practices across social media ecosystems.

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

Tether and UN join to boost digital security in Africa

Tether has joined the UN Office on Drugs and Crime to enhance cybersecurity and digital asset education across Africa. The collaboration aims to reduce vulnerabilities to cybercrime and safeguard communities against online scams and fraud.

Africa, emerging as the third-fastest-growing crypto region, faces increasing threats from digital asset fraud. A recent Interpol operation uncovered $260 million in illicit crypto and fiat across Africa, highlighting the urgent need for stronger digital security.

The partnership includes several key initiatives. In Senegal, youth will participate in a multi-phase cybersecurity education programme featuring boot camps, mentorship, and micro-grants to support innovative projects.

Civil society organisations across Africa will receive funding to support human trafficking victims in Nigeria, DRC, Malawi, Ethiopia, and Uganda. In Papua New Guinea, universities will host competitions to promote financial inclusion and prevent digital asset fraud using blockchain solutions.

Tether and UNODC aim to create secure digital ecosystems, boost economic opportunities, and equip communities to prevent organised crime. Coordinated action across sectors is considered vital to creating safer and more inclusive environments for vulnerable populations.

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

Gmail enters the Gemini era with AI-powered inbox tools

Google is reshaping Gmail around its Gemini AI models, aiming to turn email into a proactive assistant for more than three billion users worldwide.

With inbox volumes continuing to rise, the focus shifts towards managing information flows instead of simply sending and receiving messages.

New AI Overviews allow Gmail to summarise long email threads and answer natural language questions directly from inbox content.

Users can retrieve details from past conversations without complex searches, while conversation summaries roll out globally at no cost, with advanced query features reserved for paid AI subscriptions.

Writing tools are also expanding, with Help Me Write, upgraded Suggested Replies, and Proofread features designed to speed up drafting while preserving individual tone and style.

Deeper personalisation is planned through connections with other Google services, enabling emails to reflect broader user context.

A redesigned AI Inbox further prioritises urgent messages and key tasks by analysing communication patterns and relationships.

Powered by Gemini 3, these features begin rolling out in the US in English, with additional languages and regions scheduled to follow during 2026.

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

EU faces pressure to strengthen Digital Markets Act oversight

Rivals of major technology firms have criticised the European Commission for weak enforcement of the Digital Markets Act, arguing that slow procedures and limited transparency undermine the regulation’s effectiveness.

Feedback gathered during a Commission consultation highlights concerns about delaying tactics, interface designs that restrict user choice, and circumvention strategies used by designated gatekeepers.

The Digital Markets Act entered into force in March 2024, prompting several non-compliance investigations against Apple, Meta and Google. Although Apple and Meta have already faced fines, follow-up proceedings remain ongoing, while Google has yet to receive sanctions.

Smaller technology firms argue that enforcement lacks urgency, particularly in areas such as self-preferencing, data sharing, interoperability and digital advertising markets.

Concerns also extend to AI and cloud services, where respondents say the current framework fails to reflect market realities.

Generative AI tools, such as large language models, raise questions about whether existing platform categories remain adequate or whether new classifications are necessary. Cloud services face similar scrutiny, as major providers often fall below formal thresholds despite acting as critical gateways.

The Commission plans to submit a review report to the European Parliament and the Council by early May, drawing on findings from the consultation.

Proposed changes include binding timelines and interim measures aimed at strengthening enforcement and restoring confidence in the bloc’s flagship competition rules.

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

ChatGPT reaches 40 million daily users for health advice

More than 40 million people worldwide now use ChatGPT daily for health-related advice, according to OpenAI.

Over 5 percent of all messages sent to the chatbot relate to healthcare, with three in five US adults reporting use in the past three months. Many interactions occur outside clinic hours, highlighting the demand for AI guidance in navigating complex medical systems.

Users primarily turn to AI to check symptoms, understand medical terms, and explore treatment options.

OpenAI emphasises that ChatGPT helps patients gain agency over their health, particularly in rural areas where hospitals and specialised services are scarce.

The technology also supports healthcare professionals by reducing administrative burdens and providing timely information.

Despite growing adoption, regulatory oversight remains limited. Some US states have attempted to regulate AI in healthcare, and lawsuits have emerged over cases where AI-generated advice has caused harm.

OpenAI argues that ChatGPT supplements rather than replaces medical services, helping patients interpret information, prepare for care, and navigate gaps in access.

Healthcare workers are also increasingly using AI. Surveys show that two in five US professionals, including nurses and pharmacists, use generative AI weekly to draft notes, summarise research, and streamline workflows.

OpenAI plans to release healthcare policy recommendations to guide the responsible adoption of AI in clinical settings.

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

Sedgwick breach linked to TridentLocker ransomware attack

Sedgwick has confirmed a data breach at its government-focused subsidiary after the TridentLocker ransomware group claimed responsibility for stealing 3.4 gigabytes of data. The incident underscores growing threats to federal contractors handling sensitive US agency information.

The company said the breach affected only an isolated file transfer system used by Sedgwick Government Solutions, which serves agencies such as DHS, ICE, and CISA. Segmentation reportedly prevented any impact on wider corporate systems or ongoing client operations.

TridentLocker, a ransomware-as-a-service group that appeared in late 2025, listed Sedgwick Government Solutions on its dark web leak site and posted samples of stolen documents. The gang is known for double-extortion tactics, combining data encryption and public exposure threats.

Sedgwick has informed US law enforcement and affected clients while continuing to investigate with external cybersecurity experts. The firm emphasised operational continuity and noted no evidence of intrusion into its claims management servers.

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

Researchers launch AURA to protect AI knowledge graphs

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!

Healthcare systems face mounting risk from CrazyHunter ransomware

CrazyHunter ransomware has emerged as a growing threat to healthcare organisations, with repeated attacks targeting hospitals and medical service providers. The campaign focuses on critical healthcare infrastructure, raising concerns about service disruption and the exposure of sensitive patient data.

The malware is developed in Go and demonstrates a high level of technical maturity. Attackers gain initial access by exploiting weak Active Directory credentials, then use Group Policy Objects to distribute the ransomware rapidly across compromised networks.

Healthcare institutions in Taiwan have been among the most affected, with multiple confirmed incidents reported by security researchers. The pattern suggests a targeted campaign rather than opportunistic attacks, increasing pressure on regional healthcare providers to strengthen defences.

Once deployed, CrazyHunter turns off security tools and encrypts files to conceal its activity. Analysts note the use of extensive evasion techniques, including memory-based execution and redundant encryption methods, to ensure the delivery of the payload.

CrazyHunter employs a hybrid encryption scheme that combines ChaCha20 and elliptic curve cryptography, utilising partial file encryption to expedite the impact. Encrypted files receive a ‘.Hunter’ extension, with recovery dependent on the attackers’ private keys, reinforcing the pressure to pay ransoms.

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

Diplo chatbot can make mistakes. Consider checking important information.