Economic

AI reforms in Hong Kong expected to save millions in public services

Hong Kong will establish a new team to advance the use of AI across government departments, Chief Executive John Lee confirmed during his 2025 Policy Address.

The AI Efficacy Enhancement Team, led by Deputy Chief Secretary Warner Cheuk, will coordinate reforms to modernise outdated processes and promote efficiency.

Lee said his administration would focus on safe ‘AI+ development’, applying the technology in public services and encouraging adoption across different sectors instead of relying on traditional methods.

He added that Hong Kong had the potential to grow into a global hub for AI and would treat the field as a core industry for the city’s economic future.

Examples of AI adoption are already visible.

The government’s 1823 enquiry hotline uses voice recognition to cut response times by 30 per cent, while the Census and Statistics Department applies AI models to trade data and company reports, reducing manual checks by 40 per cent and improving accuracy.

Authorities expect upcoming censuses in 2026 and 2031 to save about $680 million through AI and data science technologies instead of conventional manpower-heavy methods.

The announcement comes shortly after China unveiled its national AI policy blueprint, which seeks widespread integration of the technology in research, governance and industry, with a target of 90 per cent prevalence by 2030.

Hong Kong’s approach is being positioned as part of a wider push for technological leadership in the region.

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

Microsoft seizes 338 sites tied to phishing service

Microsoft has disrupted RaccoonO365, a fast-growing phishing service used by cybercriminals to steal Microsoft 365 login details.

Using a court order from the Southern District of New York, in the US, its Digital Crimes Unit seized 338 websites linked to the operation. The takedown cut off infrastructure that enabled criminals to mimic Microsoft branding and trick victims into sharing their credentials.

Since mid-2024, RaccoonO365 has been used in at least 94 countries and has stolen more than 5,000 credentials. The kits were marketed on Telegram to hundreds of paying subscribers, including campaigns that targeted healthcare providers in the US.

Microsoft identified the group’s alleged leader as Joshua Ogundipe, based in Nigeria, who is accused of creating and promoting the service. The company has referred the case to international law enforcement while continuing efforts to dismantle any rebuilt networks.

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

WEF urges trade policy shift to protect workers in digital economy

The World Economic Forum (WEF) has published an article on using trade policy to build a fairer digital economy. Digital services now make up over half of global exports, with AI investment projected at $252 billion in 2024. Countries from Kenya to the UAE are positioning as digital hubs, but job quality still lags.

Millions of platform workers face volatile pay, lack of contracts, and no access to social protections. In Kenya alone, 1.9 million people rely on digital work yet face algorithm-driven pay systems and sudden account deactivations. India and the Philippines show similar patterns.

AI threatens to automate lower-skilled tasks such as data annotation and moderation, deepening insecurity in sectors where many developing countries have found a competitive edge. Ethical standards exist but have little impact without enforcement or supportive regulation.

Countries are experimenting with reforms: Singapore now mandates injury compensation and retirement savings for platform workers, while the Rider Law in Spain reclassifies food couriers as employees. Yet overly strict regulation risks eroding the flexibility that attracts youth and caregivers to gig work.

Trade agreements, such as the AfCFTA and the Kenya–EU pact, could embed labour protections in digital markets. Coordinated policies and tripartite dialogue are essential to ensure the digital economy delivers growth, fairness, and dignity for workers.

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

Google launches AI protocol for digital payments

Google has unveiled the Agent Payments Protocol (AP2), a new system enabling AI applications to send and receive payments, including stablecoins pegged to traditional currencies.

Developed with Coinbase, the Ethereum Foundation, and over 60 other finance and technology firms, AP2 aims to standardise transactions between AI agents and merchants.

The protocol builds on Google’s earlier Agent2Agent framework, extending it to financial interactions. AP2 supports credit and debit cards, bank transfers, and stablecoins, providing a secure and compliant foundation for automated payments.

By introducing a shared language for AI-led transactions, the system addresses risks linked to authorisation, authenticity, and accountability without human intervention.

The project reflects growing interest in stablecoins, whose circulation recently rose to $289 billion from $205 billion at the start of the year. Integrating stablecoins into AI could change how automated systems manage payments, from daily purchases to complex financial tasks.

Google and its collaborators emphasise AP2’s goal of interoperability across industries, offering flexibility, compliance, and scalability. The initiative makes digital money central to AI, signalling a shift in automated financial transactions.

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

New YouTube AI features make Shorts faster and smarter

YouTube has unveiled a new suite of AI tools designed to enhance the creation of Shorts, with its headline innovation being Veo 3 Fast, a streamlined version of Google DeepMind’s video model.

A system that can generate 480p clips with sound almost instantly, marking the first time audio has been added to Veo-generated Shorts. It is already being rolled out in the US, the UK, Canada, Australia and New Zealand, with other regions to follow instead of a limited release.

The platform also introduced several advanced editing features, such as motion transfer from video to still images, text-based styling, object insertion and Speech to Song Remixing, which converts spoken dialogue into music through DeepMind’s Lyria 2 model.

Testing will begin in the US before global expansion.

Another innovation, Edit with AI, automatically assembles raw footage into a rough cut complete with transitions, music and interactive voiceovers. YouTube confirmed the tool is in trials and will launch in select markets within weeks instead of years.

All AI-generated Shorts will display labels and watermarks to maintain transparency, as YouTube pushes to expand creator adoption and boost Shorts’ growth as a rival to TikTok and Instagram Reels.

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

Tencent launches scenario-based AI globally to boost industrial efficiency

Tencent has announced the global rollout of scenario-based AI capabilities to help enterprises accelerate industrial efficiency. At its 2025 Global Digital Ecosystem Summit, held in Shenzhen, the company introduced its Agent Development Platform 3.0 (ADP) via Tencent Cloud.

ADP enables businesses to generate autonomous AI agents that can be integrated into workflows, including customer service, marketing, inventory management, and research.

Tencent is also upgrading its internal models and infrastructure, such as ‘Agent Runtime’, to support stable, secure, and business-aligned agent deployment.

Other new tools include the SaaS+AI toolkit, which enhances productivity in office collaboration (for example, AI Minutes in Tencent Meetings) and knowledge management via Tencent LearnShare. A coding assistant called CodeBuddy is claimed to reduce developers’ coding time by 40 percent while increasing R&D efficiency by about 16 percent.

In line with its international expansion, Tencent Cloud announced that its overseas client base has doubled since last year and that it now operates across over 20 regions.

The rollout also includes open-source contributions: multilingual translation models, large multimodal models, and new Hunyuan 3D creative tools have been made available globally.

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

Swiss banks achieve milestone with deposit token test

PostFinance, Sygnum Bank and UBS, under the Swiss Bankers Association, have successfully tested a deposit token on a public blockchain. For the first time, banks completed a legally binding cross-bank payment with tokenised deposits, a breakthrough for Switzerland’s financial sector.

The proof of concept (PoC) explored how bank deposits can be represented on blockchain to enable faster, programmable and fully automated payments. Smart contracts on the system allow transactions to be triggered only when agreed conditions are met, reducing risk while increasing efficiency.

Use cases tested included payments between customers of different banks and escrow-style settlements for tokenised assets.

Results confirm that deposit tokens are technically and legally feasible across banks when combined with a public blockchain and permissioned applications. Such infrastructure could enable automated insurance settlements, instant securities trading, and machine-to-machine payments.

Industry leaders stressed that while the PoC proves feasibility, scaling will require broader cooperation with other institutions, regulators and infrastructure providers.

Project representatives described the work as a decisive step toward digital money innovation in Switzerland, reinforcing the country’s role as a pioneer in blockchain-based finance.

The Swiss Bankers Association confirmed that the project aligns with its strategic focus on digital currencies, though the results do not yet mean deposit tokens will be formally launched.

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

Hollywood studios take legal action against MiniMax for AI copyright infringement

Disney, Warner Bros. Discovery and NBCUniversal have filed a lawsuit in California against Chinese AI company MiniMax, accusing it of large-scale copyright infringement.

The studios allege that MiniMax’s Hailuo AI service generates unauthorised images and videos featuring well-known characters such as Darth Vader, marketing itself as a ‘Hollywood studio in your pocket’ instead of respecting copyright laws.

According to the complaint, MiniMax, reportedly worth $4 billion, ignored cease-and-desist requests and continues to profit from copyrighted works. The studios argue that the company could easily implement safeguards, pointing to existing controls that already block violent or explicit content.

MiniMax’s approach, as they claim, represents a serious threat to both creators and the broader film industry, which contributes hundreds of billions of dollars to the US economy.

Plaintiffs, including Disney’s Marvel and Lucasfilm units, Universal’s DreamWorks Animation and Warner Bros.’ DC Comics, are seeking statutory damages of up to $150,000 per infringed work or unspecified compensation.

They are also asking for an injunction to prevent MiniMax from continuing its alleged violations instead of simply paying damages.

The Motion Picture Association has backed the lawsuit, with its chairman Charles Rivkin warning that unchecked copyright infringement could undermine millions of jobs and the cultural value created by the American film industry.

MiniMax, based in Shanghai, has not responded publicly to the claims but has previously described itself as a global AI foundation model company with over 157 million users worldwide.

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

Prolonged JLR shutdown threatens UK export targets

Jaguar Land Rover (JLR) has confirmed that its production halt will continue until at least Wednesday, 24 September, as it works to recover from a major cyberattack that disrupted its IT systems and paralysed production at the end of August.

JLR stated that the extension was necessary because forensic investigations were ongoing and the controlled restart of operations was taking longer than anticipated. The company stressed that it was prioritising a safe and stable restart and pledged to keep staff, suppliers, and partners regularly updated.

Reports suggest recovery could take weeks, impacting production and sales channels for an extended period. Approximately 33,000 employees remain at home as factory and sales processes are not fully operational, resulting in estimated losses of £1 billion in revenue and £70 million in profits.

The shutdown also poses risks to the wider UK economy, as JLR represents roughly four percent of British exports. The incident has renewed calls for the Cyber Security and Resilience Bill, which aims to strengthen defenses against digital threats to critical industries.

No official attribution has been made, but a group calling itself Scattered Lapsus$ Hunters has claimed responsibility. The group claims to have deployed ransomware and published screenshots of JLR’s internal SAP system, linking itself to extortion groups, including Scattered Spider, Lapsus$, and ShinyHunters.

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

Cyberattack compromises personal data used for DBS checks at UK college

Bracknell and Wokingham College has confirmed a cyberattack that compromised data collected for Disclosure and Barring Service (DBS) checks. The breach affects data used by Activate Learning and other institutions, including names, dates of birth, National Insurance numbers, and passport details.

Access Personal Checking Services (APCS) was alerted by supplier Intradev on August 17 that its systems had been accessed without authorisation. While payment card details and criminal conviction records were not compromised, data submitted between December 2024 and May 8, 2025, was copied.

APCS stated that its own networks and those of Activate Learning were not breached. The organisation is contacting only those data controllers where confirmed breaches have occurred and has advised that its services can continue to be used safely.

Activate Learning reported the incident to the Information Commissioner’s Office following a risk assessment. APCS is still investigating the full scope of the breach and has pledged to keep affected institutions and individuals informed as more information becomes available.

Individuals have been advised to closely monitor their financial statements, exercise caution when opening phishing emails, and regularly update security measures, including passwords and two-factor authentication. Activate Learning emphasised the importance of staying vigilant to minimise risks.

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