Competition regulation

Dutch watchdog warns AI chatbots threaten election integrity

The Dutch data authority warns AI chatbots are biased and unreliable for voting advice ahead of national elections. An AP investigation found chatbots often steered users to the same two parties, ignoring their actual preferences.

In over half of the tests, the bots suggested either Geert Wilders’ far-right Freedom Party (PVV) or the leftwing GroenLinks-PvdA led by Frans Timmermans. Other parties, such as the centre-right CDA, were rarely mentioned even when users’ answers closely matched their platforms.

AP deputy head Monique Verdier said that voters were being steered towards parties that did not necessarily reflect their political views, warning that this undermines the integrity of free and fair elections.

The report comes ahead of the 29 October election, where the PVV currently leads the polls. However, the race remains tight, with GroenLinks-PvdA and CDA still in contention and many voters undecided.

Although the AP noted that the bias was not intentional, it attributed the problem to the way AI chatbots function, highlighting the risks of relying on opaque systems for democratic decisions.

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

ChatGPT to exit WhatsApp after Meta policy change

OpenAI says ChatGPT will leave WhatsApp on 15 January 2026 after Meta’s new rules banning general-purpose AI chatbots on the platform. ChatGPT will remain available on iOS, Android, and the web, the company said.

Users are urged to link their WhatsApp number to a ChatGPT account to preserve history, as WhatsApp doesn’t support chat exports. OpenAI will also let users unlink their phone numbers after linking.

Until now, users could message ChatGPT on WhatsApp to ask questions, search the web, generate images, or talk to the assistant. Similar third-party bots offered comparable features.

Meta quietly updated WhatsApp’s business API to prohibit AI providers from accessing or using it, directly or indirectly. The change effectively forces ChatGPT, Perplexity, Luzia, Poke, and others to shut down their WhatsApp bots.

The move highlights platform risk for AI assistants and shifts demand toward native apps and web. Businesses relying on WhatsApp AI automations will need alternatives that comply with Meta’s policies.

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

Innovation versus risk shapes Australia’s AI debate

Australia’s business leaders were urged to adopt AI now to stay competitive, despite the absence of hard rules, at the AI Leadership Summit in Brisbane. The National AI Centre unveiled revised voluntary guidelines, and Assistant Minister Andrew Charlton said a national AI plan will arrive later this year.

The guidance sets six priorities, from stress-testing and human oversight to clearer accountability, aiming to give boards practical guardrails. Speakers from NVIDIA, OpenAI, and legal and academic circles welcomed direction but pressed for certainty to unlock stalled investment.

Charlton said the plan will focus on economic opportunity, equitable access, and risk mitigation, noting some harms are already banned, including ‘nudify’ apps. He argued Australia will be poorer if it hesitates, and regulators must be ready to address new threats directly.

The debate centred on proportional regulation: too many rules could stifle innovation, said Clayton Utz partner Simon Newcomb, yet delays and ambiguity can also chill projects. A ‘gap analysis’ announced by Treasurer Jim Chalmers will map which risks existing laws already cover.

CyberCX’s Alastair MacGibbon warned that criminals are using AI to deliver sharper phishing attacks and flagged the return of erotic features in some chatbots as an oversight test. His message echoed across panels: move fast with governance, or risk ceding both competitiveness and safety.

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

AI-generated images used in jewellery scam

A jeweller in Hove is dealing with daily complaints from customers of a similarly named but fraudulent business. Stevie Holmes runs Scarlett Jewellery but keeps receiving complaints from customers who confused it with the AI-driven Scarlett Jewels website.

Many reported receiving poor-quality goods or nothing at all.

Holmes said the mix-ups have kept her occupied for at least an hour a day since July. Without clarification, people could post negative comments about her genuine business on social media, potentially damaging its reputation.

Scarlett Jewels is run by Denimtex Limited with an address in Hong Kong, though its website claims a personal story of a retiring designer.

Experts say such scams are increasingly common due to how easy and cheap it is to create AI images. Professor Ana Canhoto from the University of Sussex noted AI-generated product photos often appear too perfect or flawed, while fake reviews and claims of scarcity are typical tactics to mislead buyers.

Trustpilot ratings for Scarlett Jewels are mostly one star, with customers describing items as ‘tat’ or ‘poor quality’.

Authorities are taking action, with the Advertising Standards Authority banning similar ads and Facebook restricting Scarlett Jewels from creating new adverts. Buyers are advised to spot off AI images, large discounts, and genuine reviews to avoid falling for scams.

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

Tailored pricing is here and personal data is the price signal

AI is quietly changing how prices are set online. Beyond demand-based shifts, companies increasingly tailor offers to individuals, using browsing history, purchase habits, device, and location to predict willingness to pay. Two shoppers may see different prices for the same product at the same moment.

Dynamic pricing raises or lowers prices for everyone as conditions change, such as school-holiday airfares or hotel rates during major events. Personalised pricing goes further by shaping offers for specific users, rewarding cart-abandoners with discounts while charging rarer shoppers a premium.

Platforms mine clicks, time on page, past purchases, and abandoned baskets to build profiles. Experiments show targeted discounts can lift sales while capping promo spend, proving engineered prices scale. The result: you may not see a ‘standard’ price, but one designed for you.

The risks are mounting. Income proxies such as postcode or device can entrench inequality, while hidden algorithms erode trust when buyers later find cheaper prices. Accountability is murky if tailored prices mislead, discriminate, or breach consumer protections without clear disclosure.

Regulators are moving. A competition watchdog in Australia has flagged transparency gaps, unfair trading risks, and the need for algorithmic disclosure. Businesses now face a twin test: deploy AI pricing with consent, explainability, and opt-outs, and prove it delivers value without crossing ethical lines.

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

AWS glitch triggers widespread outages across major apps

A major internet outage hit some of the world’s biggest apps and sites from about 9 a.m. CET Monday, with issues traced to Amazon Web Services. Tracking sites reported widespread failures across the US and beyond, disrupting consumer and enterprise services.

AWS cited ‘significant error rates’ in DynamoDB requests in the US-EAST-1 region, impacting additional services in Northern Virginia. Engineers are mitigating while investigating root cause, and some customers couldn’t create or update Support Cases.

Outages clustered around Virginia’s dense data-centre corridor but rippled globally. Impacted brands included Amazon, Google, Snapchat, Roblox, Fortnite, Canva, Coinbase, Slack, Signal, Vodafone and the UK tax authority HMRC.

Coinbase told users ‘all funds are safe’ as platforms struggled to authenticate, fetch data and serve content tied to affected back-ends. Third-party monitors noted elevated failure rates across APIs and app logins.

The incident underscores heavy reliance on hyperscale infrastructure and the blast radius when core data services falter. Full restoration and a formal post-mortem are pending from AWS.

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

Labels and Spotify align on artist-first AI safeguards

Spotify partners with major labels on artist-first AI tools, putting consent and copyright at the centre of product design. The plan aims to align new features with transparent labelling and fair compensation while addressing concerns about generative music flooding platforms.

The collaboration with Sony, Universal, Warner, and Merlin will give artists control over participation in AI experiences and how their catalogues are used. Spotify says it will prioritise consent, clearer attribution, and rights management as it builds new tools.

Early direction points to expanded labelling via DDEX, stricter controls against mass AI uploads, and protections against search and recommendation manipulation. Spotify’s AI DJ and prompt-based playlists hint at how engagement features could evolve without sidelining creators.

Future products are expected to let artists opt in, monitor usage, and manage when their music feeds AI-generated works. Rights holders and distributors would gain better tracking and payment flows as transparency improves across the ecosystem.

Industry observers say the tie-up could set a benchmark for responsible AI in music if enforcement matches ambition. By moving in step with labels, Spotify is pitching a path where innovation and artist advocacy reinforce rather than undermine each other.

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

Between trips, Uber pilots paid AI data work

Uber is piloting ‘Digital Tasks’ in the US, letting select drivers and couriers earn by training AI models between trips.

Tasks include short selfie videos in any language, uploading multilingual documents, and uploading category-tagged images; each takes minutes, and pay varies by task.

Uber says demand came from US drivers seeking off-road income; participants can opt in via the Work Hub and need no extra experience.

Partners commissioning the data aren’t named. The pilot starts later this year, with potential expansion to non-drivers and wider markets.

The move diversifies beyond rides and delivery as robotaxis loom. Uber argues for more earning channels now, while autonomy scales over time.

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

Lehane backs OpenAI’s Australia presence as AI copyright debate heats up

OpenAI signalled a break with Australia’s tech lobby on copyright, with global affairs chief Chris Lehane telling SXSW Sydney the company’s models are ‘going to be in Australia, one way or the other’, regardless of reforms or data-mining exemptions.

Lehane framed two global approaches: US-style fair use that enables ‘frontier’ AI, versus a tighter, historical copyright that narrows scope, saying OpenAI will work under either regime. Asked if Australia risked losing datacentres without loser laws, he replied ‘No’.

Pressed on launching and monetising Sora 2 before copyright issues are settled, Lehane argued innovation precedes adaptation and said OpenAI aims to ‘benefit everyone’. The company paused videos featuring Martin Luther King Jr.’s likeness after family complaints.

Lehane described the US-China AI rivalry as a ‘very real competition’ over values, predicting that one ecosystem will become the default. He said US-led frontier models would reflect democratic norms, while China’s would ‘probably’ align with autocratic ones.

To sustain a ‘democratic lead’, Lehane said allies must add gigawatt-scale power capacity each week to build AI infrastructure. He called Australia uniquely positioned, citing high AI usage, a 30,000-strong developer base, fibre links to Asia, Five Eyes membership, and fast-growing renewables.

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

Microsoft to support UAE investment analytics with responsible AI tools

The UAE Ministry of Investment and Microsoft signed a Memorandum of Understanding at GITEX Global 2025 to apply AI to investment analytics, financial forecasting, and retail optimisation. The deal aims to strengthen data governance across the investment ecosystem.

Under the MoU, Microsoft will support upskilling through its AI National Skilling Initiative, targeting 100,000 government employees. Training will focus on practical adoption, responsible use, and measurable outcomes, in line with the UAE’s National AI Strategy 2031.

Both parties will promote best practices in data management using Azure services such as Data Catalog and Purview. Workshops and knowledge-sharing sessions with local experts will standardise governance. Strong controls are positioned as the foundation for trustworthy AI at scale.

The agreement was signed by His Excellency Mohammad Alhawi and Amr Kamel. Officials say the collaboration will embed AI agents into workflows while maintaining compliance. Investment teams are expected to gain real-time insights and automation that shorten the time to action.

The partnership supports the ambition to make the UAE a leader in AI-enabled investment. It also signals deeper public–private collaboration on sovereign capabilities. With skills, standards, and use cases in place, the ministry aims to attract capital and accelerate diversification.

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