Consumer protection

Orange suffers major data breach

Orange Belgium has confirmed a data breach affecting 850,000 customers, after a cyberattack targeted one of its internal IT systems. The attack, discovered in late July, exposed names, phone numbers, SIM card details, tariff plans and PUK codes. No financial or password data was compromised.

The telecoms provider blocked access to the affected system and notified authorities. A formal complaint has also been filed with the judiciary. All affected users are being informed via email or SMS and are urged to stay alert for phishing and identity fraud attempts.

Orange Belgium has advised users to strengthen account security with strong, unique passwords and to be cautious of suspicious links and messages. This marks the third cyber incident involving Orange in 2025, following earlier attacks, though those breaches varied in impact.

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

Australia expands crackdown on online scams

Australia has taken down 14,000 online scams since July 2023, with more than 3,000 involving crypto. The Australian Securities and Investments Commission (ASIC) has expanded scam enforcement to cover social media ads, investment fraud, and phishing websites.

ASIC Deputy Chair Sarah Court noted takedown powers refer suspicious sites to cybercrime specialists for removal. Common scams include AI trading bots, fake websites, and fraudulent celebrity endorsements, making fraud harder to detect.

Investment scams remain the leading threat, with over $73 million lost this year, though overall losses have fallen since 2023. Regulators urged caution with testimonials, AI investment claims, and schemes on WhatsApp, Telegram, and other messaging apps.

Crypto ATMs have also come under scrutiny. AUSTRAC and the AFP have investigated connections between crypto ATMs and scams, including pig-butchering operations. Australia has nearly 2,000 crypto ATMs, with new limits to curb crime and protect investors.

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

Google enhances AI Mode with personalised dining suggestions

Google has expanded its AI Mode in Search to 180 additional countries and territories, introducing new agentic features to help users make restaurant reservations. The service remains limited to English and is not yet available in the European Union.

The update enables users to specify their dining preferences and constraints, allowing the system to scan multiple platforms and present real-time availability. Once a choice is made, users are directed to the restaurant’s booking page.

Partners supporting the service include OpenTable, Resy, SeatGeek, StubHub, Booksy, Tock, and Ticketmaster. The feature is part of Google’s Search Labs experiment, available to subscribers of Google AI Ultra in the United States.

AI Mode also tailors suggestions based on previous searches and introduces a Share function, letting users share restaurant options or planning results with others, with the option to delete links.

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

Court filing details Musk’s outreach to Zuckerberg over OpenAI bid

Elon Musk attempted to bring Meta chief executive Mark Zuckerberg into his consortium’s $97.4 billion bid for OpenAI earlier this year, the company disclosed in a court filing.

According to sworn interrogations, OpenAI said Musk had discussed possible financing arrangements with Zuckerberg as part of the bid. Musk’s AI startup xAI, a competitor to OpenAI, did not respond to requests for comment.

In the filing, OpenAI asked a federal judge to order Meta to provide documents related to any bid for OpenAI, including internal communications about restructuring or recapitalisation. The firm argued these records could clarify motivations behind the bid.

Meta countered that such documents were irrelevant and suggested OpenAI seek them directly from Musk or xAI. A US judge ruled that Musk must face OpenAI’s claims of attempting to harm the company through public remarks and what it described as a sham takeover attempt.

The legal dispute follows Musk’s lawsuit against OpenAI and Sam Altman over its for-profit transition, with OpenAI filing a countersuit in April. A jury trial is scheduled for spring 2026.

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

Ransomware attack at DaVita exposes data of 2.7 million patients in the US

A ransomware attack against dialysis provider DaVita has exposed the personal data of 2.7 million people, according to a notice on the US health department’s website.

The company first disclosed the cyber incident in April, saying it had taken steps to restore operations but could not predict the scale of disruption.

DaVita confirmed that hackers gained unauthorised access to its laboratory database, which contained sensitive information belonging to some current and former patients. The firm said it is now contacting those affected and offering free credit monitoring to help protect against identity theft.

Despite the intrusion, DaVita maintained uninterrupted dialysis services across its network of nearly 3,000 outpatient clinics and home treatment programmes. The company described the cyberattack as a temporary disruption but stressed that patient care was never compromised.

Financial disclosures show the incident led to around $13.5 million in charges during the second quarter of 2025. Most of the costs were linked to system restoration and third-party support, with $1 million attributed to higher patient care expenses.

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

Google pushes agentic AI worldwide with AI Mode rollout

Google has expanded its AI Mode service to 180 additional countries, extending advanced agentic capabilities to a global audience.

Previously available only in markets such as the US, UK and India, the service allows users to search for information and carry out tasks on their behalf. The update reflects Google’s ambition to move from simple answers to action-oriented assistance.

A key rollout feature is the restaurant booking tool for AI Ultra subscribers. Using natural language requests such as ”find a romantic Italian spot for two tonight,” the system can check availability, offer personalised suggestions and confirm reservations directly within search.

The feature relies on real-time data from partners like OpenTable and highlights how Google’s AI can execute tasks instead of simply presenting options.

Further tools are expected soon, including ticketing for events and appointment scheduling. These are powered by the Gemini models, which tailor recommendations based on user behaviour and allow group planning through shared responses.

While the services could reduce reliance on third-party apps in sectors such as travel and hospitality, they also raise concerns over data privacy, inclusivity and cultural differences in an English-only rollout.

The global expansion strengthens Google’s position against rivals like Microsoft and OpenAI, who are also pushing forward in agentic AI. The company sees subscription upgrades to AI Ultra as a way to offset slower advertising growth, while early reports suggest increased user engagement.

However, the long-term impact will depend on balancing innovation with ethical safeguards as Google works to deliver more multilingual and accessible features.

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

Students seek emotional support from AI chatbots

College students are increasingly turning to AI chatbots for emotional support, prompting concern among mental health professionals. A 2025 report ranked ‘therapy and companionship’ as the top use case for generative AI, particularly among younger users.

Studies by MIT and OpenAI show that frequent AI use can lower social confidence and increase avoidance of face-to-face interaction. On campuses, digital mental health platforms now supplement counselling services, offering tools that identify at-risk students and provide basic support.

Experts warn that chatbot companionship may create emotional habits that lack grounding in reality and hinder social skill development. Counsellors advocate for educating students on safe AI use and suggest universities adopt tools that flag risky engagement patterns.

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

Europol warns that the $50,000 Qilin reward is fake

Europol has warned that a reported $50,000 reward for information on two members of the Qilin ransomware group is fake. The message, circulating on Telegram, claimed the suspects, known as Haise and XORacle, coordinate affiliates and manage extortion operations.

Europol clarified that it does not operate a Telegram channel and that the message does not originate from its official accounts, which are active on Instagram, LinkedIn, X, Bluesky, YouTube, and Facebook.

Qilin, also known as Agenda, has been active since 2022 and, in 2025, listed over 400 victims on its leak website, including media and pharmaceutical companies.

Recent attacks, such as the one targeting Inotiv, demonstrate the group’s ongoing threat. Analysts note that cybercriminals often circulate false claims to undermine competitors, mislead affiliates, or sow distrust within rival gangs.

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

Caution rises over inflated AI-driven tech valuations

US tech stocks have stumbled after a sharp rally, with investors increasingly cautious over AI-linked valuations and shifting market conditions. The S&P 500 tech sector has dropped around 2.5% this week, while the Nasdaq has slipped 2%, led by losses in Nvidia and Palantir.

The fall follows a 50% surge in tech shares since April, far outpacing the broader market and pushing valuations to year-highs. Concerns are growing that investor enthusiasm around AI has become overheated, with some funds reducing their exposure ahead of expected interest rate guidance.

US market watchers are now focused on Federal Reserve Chair Jerome Powell’s speech at Jackson Hole, which could signal if rate cuts are on the horizon. Tech stocks, already heavily weighted in many portfolios, are particularly vulnerable to higher rates due to their stretched price-to-earnings ratios.

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

AI improves customer experience at Citi

Citi has expanded its digital client platform, CitiDirect Commercial Banking, with new AI capabilities to improve customer service and security.

The platform now supports over half of Citi’s global commercial banking client base and handles around 2.3 million sessions.

AI features assist in fraud detection, automate customer queries, and provide real-time onboarding updates and guidance.

KYC renewals have been simplified through automated alerts and pre-filled forms, cutting effort and processing time for clients.

Live in markets including the UK, US, India, and others, the platform has received positive feedback from over 10,000 users. Citi says the enhancements are part of a broader effort to make mid-sized corporate banking faster, more innovative, and more efficient.

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