Cybersecurity

Digital fraud declines in Russia after rollout of Cyberbez measures

Russia has reported a sharp decline in cyber fraud following the introduction of new regulatory measures in 2025. Officials say legislative action targeting telephone and online scams has begun to deliver measurable results.

State Secretary and Deputy Minister of Digital Development Ivan Lebedev told the State Duma that crimes covered by the first package of reforms, known as ‘Cyberbez 1.0’, have fallen by 40%, according to confirmed statistics.

Earlier this year, Lebedev said Russia records roughly 677,000 cases of phone and online fraud annually, with incidents rising by more than 35% since 2022, highlighting the scale of the challenge faced by authorities.

In April, President Vladimir Putin signed a law introducing a range of countermeasures, including a state information system to combat fraud, limits on unsolicited marketing calls, stricter SIM card issuance rules, and new compliance obligations for banks.

Further steps are now under discussion. Officials say a second package is being prepared, while a third set of initiatives was announced in December as Russia continues to strengthen its digital security framework.

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

UK report quantifies rapid advances in frontier AI capabilities

For the first time, the UK has published a detailed, evidence-based assessment of frontier AI capabilities. The Frontier AI Trends Report draws on two years of structured testing across areas including cybersecurity, software engineering, chemistry, and biology.

The findings show rapid progress in technical performance. Success rates on apprentice-level cyber tasks rose from under 9% in 2023 to around 50% in 2025, while models also completed expert-level cyber challenges previously requiring a decade of experience.

Safeguards designed to limit misuse are also improving, according to the report. Red-team testing found that the time required to identify universal jailbreaks increased from minutes to several hours between model generations, representing an estimated forty-fold improvement in resistance.

The analysis highlights advances beyond cybersecurity. AI systems now complete hour-long software engineering tasks more than 40% of the time, while biology and chemistry models outperform PhD-level researchers in controlled knowledge tests and support non-experts in laboratory-style workflows.

While the report avoids policy recommendations, UK officials say it strengthens transparency around advanced AI systems. The government plans to continue investing in evaluation science through the AI Security Institute, supporting independent testing and international collaboration.

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

New Kimwolf Android botnet linked to a record-breaking DDoS attacks

Cybersecurity researchers have uncovered a rapidly expanding Android botnet known as Kimwolf, which has already compromised approximately 1.8 million devices worldwide.

The malware primarily targets smart TVs, set-top boxes, and tablets connected to residential networks, with infections concentrated in countries including Brazil, India, the US, Argentina, South Africa, and the Philippines.

Analysis by QiAnXin XLab indicates that Kimwolf demonstrates a high degree of operational resilience.

Despite multiple disruptions to its command-and-control infrastructure, the botnet has repeatedly re-emerged with enhanced capabilities, including the adoption of Ethereum Name Service to harden its communications against takedown efforts.

Researchers also identified significant similarities between Kimwolf and AISURU, one of the most powerful botnets observed in recent years. Shared source code, infrastructure, and infection scripts suggest both botnets are operated by the same threat group and have coexisted on large numbers of infected devices.

AISURU has previously drawn attention for launching record-setting distributed denial-of-service attacks, including traffic peaks approaching 30 terabits per second.

The emergence of Kimwolf alongside such activity highlights the growing scale and sophistication of botnet-driven cyber threats targeting global internet infrastructure.

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

PwC automates AI governance with Agent Mode

The global professional services network, PwC, has expanded its Model Edge platform with the launch of Agent Mode, an AI assistant designed to automate governance, compliance and documentation across enterprise AI model lifecycles.

The capability targets the growing administrative burden faced by organisations as AI model portfolios scale and regulatory expectations intensify.

Agent Mode allows users to describe governance tasks in natural language, instead of manually navigating workflows.

A system that executes actions directly within Model Edge, generates leadership-ready documentation and supports common document and reporting formats, significantly reducing routine compliance effort.

PwC estimates weekly time savings of between 20 and 50 percent for governance and model risk teams.

Behind the interface, a secure orchestration engine interprets user intent, verifies role based permissions and selects appropriate large language models based on task complexity. The design ensures governance guardrails remain intact while enabling faster and more consistent oversight.

PwC positions Agent Mode as a step towards fully automated, agent-driven AI governance, enabling organisations to focus expert attention on risk assessment and regulatory judgement instead of process management as enterprise AI adoption accelerates.

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

AI and security trends shape the internet in 2025

Cloudflare released its sixth annual Year in Review, providing a comprehensive snapshot of global Internet trends in 2025. The report highlights rising digital reliance, AI progress, and evolving security threats across Cloudflare’s network and Radar data.

Global Internet traffic rose 19 percent year-on-year, reflecting increased use for personal and professional activities. A key trend was the move from large-scale AI training to continuous AI inference, alongside rapid growth in generative AI platforms.

Google and Meta remained the most popular services, while ChatGPT led in generative AI usage.

Cybersecurity remained a critical concern. Post-quantum encryption now protects 52 percent of Internet traffic, yet record-breaking DDoS attacks underscored rising cyber risks.

Civil society and non-profit organisations were the most targeted sectors for the first time, while government actions caused nearly half of the major Internet outages.

Connectivity varied by region, with Europe leading in speed and quality and Spain ranking highest globally. The report outlines 2025’s Internet challenges and progress, providing insights for governments, businesses, and users aiming for greater resilience and security.

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

Crypto theft soars in 2025 with fewer but bigger attacks

Cryptocurrency theft intensified in 2025, with total stolen funds exceeding $3.4 billion despite fewer large-scale incidents. Losses became increasingly concentrated, with a few major breaches driving most of the annual damage and widening the gap between typical hacks and extreme outliers.

North Korea remained the dominant threat actor, stealing at least $2.02 billion in digital assets during the year, a 51% increase compared with 2024.

Larger thefts were achieved through fewer operations, often relying on insider access, executive impersonation, and long-term infiltration of crypto firms rather than frequent attacks.

Laundering activity linked to North Korean actors followed a distinctive and disciplined pattern. Stolen funds moved in smaller tranches through Chinese-language laundering networks, bridges, and mixing services, usually following a structured 45-day cycle.

Individual wallet attacks surged, impacting tens of thousands of victims, while the total value stolen from personal wallets fell. Decentralised finance remained resilient, with hack losses low despite rising locked capital, indicating stronger security practices.

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

Healthcare faces growing compliance pressure from AI adoption

AI is becoming a practical tool across healthcare as providers face rising patient demand, chronic disease and limited resources.

These AI systems increasingly support tasks such as clinical documentation, billing, diagnostics and personalised treatment instead of relying solely on manual processes, allowing clinicians to focus more directly on patient care.

At the same time, AI introduces significant compliance and safety risks. Algorithmic bias, opaque decision-making, and outdated training data can affect clinical outcomes, raising questions about accountability when errors occur.

Regulators are signalling that healthcare organisations cannot delegate responsibility to automated systems and must retain meaningful human oversight over AI-assisted decisions.

Regulatory exposure spans federal and state frameworks, including HIPAA privacy rules, FDA oversight of AI-enabled medical devices and enforcement under the False Claims Act.

Healthcare providers are expected to implement robust procurement checks, continuous monitoring, governance structures and patient consent practices as AI regulation evolves towards a more coordinated national approach.

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

US platforms signal political shift in DSA risk reports

Major online platforms have submitted their 2025 systemic risk assessments under the Digital Services Act as the European Commission moves towards issuing its first fine against a Very Large Online Platform.

The reports arrive amid mounting political friction between Brussels and Washington, placing platform compliance under heightened scrutiny on both regulatory and geopolitical fronts.

Several US-based companies adjusted how risks related to hate speech, misinformation and diversity are framed, reflecting political changes in the US while maintaining formal alignment with EU law.

Meta softened enforcement language, reclassified hate speech under broader categories and reduced visibility of civil rights structures, while continuing to emphasise freedom of expression as a guiding principle.

Google and YouTube similarly narrowed references to misinformation, replaced established terminology with less charged language and limited enforcement narratives to cases involving severe harm.

LinkedIn followed comparable patterns, removing references to earlier commitments on health misinformation, civic integrity and EU voluntary codes that have since been integrated into the DSA framework.

X largely retained its prior approach, although its report continues to reference cooperation with governments and civil society that contrasts with the platform’s public positioning.

TikTok diverged from other platforms by expanding disclosures on hate speech, election integrity and fact-checking, likely reflecting its vulnerability to regulatory action in both the EU and the US.

European regulators are expected to assess whether these shifts represent genuine risk mitigation or strategic alignment with US political priorities.

As systemic risk reports increasingly inform enforcement decisions, subtle changes in language, scope and emphasis may carry regulatory consequences well beyond their formal compliance function.

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

DDoS attacks in 2025 become faster and smarter

DDoS attacks in 2025 became short and automated, often ending in minutes with minimal warning. Traditional response times are now insufficient against these high-speed threats.

Attackers increasingly use multiple hosts and blended vectors, including TCP, UDP, DNS, and SYN floods. IoT botnets and residential proxies amplify scale, with global capacity exceeding 250 Tbps.

Algorithmic orchestration allows attacks to adapt and escalate automatically. Even low-tech campaigns remain disruptive to weaker networks, highlighting the need for continuous monitoring.

Defenders must adopt AI-driven, sub-minute mitigation and self-defending architectures. Real-time detection is now essential to maintain uptime and prevent reputational damage.

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

New 5G-advanced upgrade boosts UAE connectivity

UAE telecom operator ‘du’ has deployed the country’s first tri-band Radio Remote Unit on the 600MHz spectrum in partnership with Huawei. The rollout marks progress in the UAE’s 5G-Advanced network development.

Improved indoor coverage and faster speeds are delivered through dynamic power sharing and multi-band functionality. The upgrade supports services such as 5G Fixed Wireless Access and Voice over New Radio.

Lower energy consumption and a compact design reduce the environmental footprint of network infrastructure. The deployment aligns with national sustainability goals while improving long-term network efficiency.

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