Content Policy

AI bot swarms emerge as a new threat to democracy

Researchers and free-speech advocates are warning that coordinated swarms of AI agents could soon be deployed to manipulate public opinion at a scale capable of undermining democratic systems.

According to a consortium of academics from leading universities, advances in generative and agentic AI now enable large numbers of human-like bots to infiltrate online communities and autonomously simulate organic political discourse.

Unlike earlier forms of automated misinformation, AI swarms are designed to adapt to social dynamics, learn community norms and exchange information in pursuit of a shared objective.

By mimicking human behaviour and spreading tailored narratives gradually, such systems could fabricate consensus, amplify doubt around electoral processes and normalise anti-democratic outcomes without triggering immediate detection.

Evidence of early influence operations has already emerged in recent elections across Asia, where AI-driven accounts have engaged users with large volumes of unverifiable information rather than overt propaganda.

Researchers warn that information overload, strategic neutrality and algorithmic amplification may prove more effective than traditional disinformation campaigns.

The authors argue that democratic resilience now depends on global coordination, combining technical safeguards such as watermarking and detection tools with stronger governance of political AI use.

Without collective action, they caution that AI-enabled manipulation risks outpacing existing regulatory and institutional defences.

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

Grok faces regulatory scrutiny in South Korea over explicit AI content

South Korea has moved towards regulatory action against Grok, the generative AI chatbot developed by xAI, following allegations that the system was used to generate and distribute sexually exploitative deepfake images.

The country’s Personal Information Protection Commission has launched a preliminary fact-finding review to assess whether violations occurred and whether the matter falls within its legal remit.

The review follows international reports accusing Grok of facilitating the creation of explicit and non-consensual images of real individuals, including minors.

Under the Personal Information Protection Act of South Korea, generating or altering sexual images of identifiable people without consent may constitute unlawful handling of personal data, exposing providers to enforcement action.

Concerns have intensified after civil society groups estimated that millions of explicit images were produced through Grok over a short period, with thousands involving children.

Several governments, including those in the US, Europe and Canada, have opened inquiries, while parts of Southeast Asia have opted to block access to the service altogether.

In response, xAI has introduced technical restrictions preventing users from generating or editing images of real people. Korean regulators have also demanded stronger youth protection measures from X, warning that failure to address criminal content involving minors could result in administrative penalties.

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

France accelerates rapid ban on social media for under-15s

French President Emmanuel Macron has called for an accelerated legislative process to introduce a nationwide ban on social media for children under 15 by September.

Speaking in a televised address, Macron said the proposal would move rapidly through parliament so that explicit rules are in place before the new school year begins.

Macron framed the initiative as a matter of child protection and digital sovereignty, arguing that foreign platforms or algorithmic incentives should not shape young people’s cognitive and emotional development.

He linked excessive social media use to manipulation, commercial exploitation and growing psychological harm among teenagers.

Data from France’s health watchdog show that almost half of teenagers spend between two and five hours a day on their smartphones, with the vast majority accessing social networks daily.

Regulators have associated such patterns with reduced self-esteem and increased exposure to content linked to self-harm, drug use and suicide, prompting legal action by families against major platforms.

The proposal from France follows similar debates in the UK and Australia, where age-based access restrictions have already been introduced.

The French government argues that decisive national action is necessary instead of waiting for a slower Europe-wide consensus, although Macron has reiterated support for a broader EU approach.

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

New AI startup secures $5M to transform children’s digital learning

AI education start-up Sparkli has raised $5 million in seed funding to develop an ‘anti-chatbot’ AI platform to transform how children engage with digital content.

Unlike traditional chatbots that focus on general conversation, Sparkli positions its AI as an interactive learning companion, guiding kids through topics such as math, science and language skills in a dynamic, age-appropriate format.

The funding will support product development, content creation and expansion into new markets. Founders say the platform addresses increasing concerns about passive screen time by offering educational interactions that blend AI responsiveness with curriculum-aligned activities.

The company emphasises safe design and parental controls to ensure technology supports learning outcomes rather than distraction.

Investors backing Sparkli see demand for responsible AI applications for children that can enhance cognition and motivation while preserving digital well-being. As schools and homes increasingly integrate AI tools, Sparkli aims to position itself at the intersection of educational technology and child-centred innovation.

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

Writers challenge troubling AI assumptions about language and style

A growing unease among writers is emerging as AI tools reshape how language is produced and perceived. Long-established habits, including the use of em dashes and semicolons, are increasingly being viewed with suspicion as machine-generated text becomes more common.

The concern is not opposition to AI itself, but the blurring of boundaries between human expression and automated output. Writers whose work was used to train large language models without consent say stylistic traits developed over decades are now being misread as algorithmic authorship.

Academic and editorial norms are also shifting under this pressure. Teaching practices that once valued rhythm, voice, and individual cadence are increasingly challenged by stricter stylistic rules, sometimes framed as safeguards against sloppy or machine-like writing rather than as matters of taste or craft.

At the same time, productivity tools embedded into mainstream software continue to intervene in the writing process, offering substitutions and revisions that prioritise clarity and efficiency over nuance. Such interventions risk flattening language and discouraging the idiosyncrasies that define human authorship.

As AI becomes embedded in publishing, education, and professional writing, the debate is shifting from detection to preservation. Many writers warn that protecting human voice and stylistic diversity is essential, arguing that affectless, uniform prose would erode creativity and trust.

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

Hollywood figures back anti-AI campaign

More than 800 creatives in the US have signed an anti-AI campaign accusing big technology companies of exploiting human work. High-profile figures from film and television in the country have backed the initiative, which argues that training AI on creative content without consent amounts to theft.

The campaign was launched by the Human Artistry Campaign, a coalition representing creators, unions and industry groups in the country. Supporters say AI systems should not be allowed to use artistic work without permission and fair compensation.

Actors and filmmakers in the US warned that unchecked AI adoption threatens livelihoods across film, television and music. Campaign organisers said innovation should not come at the expense of creators’ rights or ownership of their work.

The statement adds to growing pressure on lawmakers and technology firms in the US. Creative workers are calling for clearer rules on how AI can be developed and deployed across the entertainment industry.

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

Analysis reveals Grok generated 3 million sexualised images

A new analysis found Grok generated an estimated three million sexualised images in 11 days, including around 23,000 appearing to depict children. The findings raise serious concerns over safeguards, content moderation, and platform responsibility.

The surge followed the launch of Grok’s one-click image editing feature in late December, which quickly gained traction among users. Restrictions were later introduced, including paid access limits and technical measures to prevent image undressing.

Researchers based their estimates on a random sample of 20,000 images, extrapolating from these results to more than 4.6 million images generated during the study period. Automated tools and manual review identified sexualised content and confirmed cases involving individuals appearing under 18.

Campaigners have warned that the findings expose significant gaps in AI safety controls, particularly in protecting children. Calls are growing for stricter oversight, stronger accountability, and more robust safeguards before large-scale AI image deployment.

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

Japan arrests suspect over AI deepfake pornography

Police in Japan have arrested a man accused of creating and selling non-consensual deepfake pornography using AI tools. The Tokyo Metropolitan Police Department said thousands of manipulated images of female celebrities were distributed through paid websites.

Investigators in Japan allege the suspect generated hundreds of thousands of images over two years using freely available generative AI software. Authorities say the content was promoted on social media before being sold via subscription platforms.

The arrest follows earlier cases in Japan and reflects growing concern among police worldwide. In South Korea, law enforcement has reported hundreds of arrests linked to deepfake sexual crimes, while cases have also emerged in the UK.

European agencies, including Europol, have also coordinated arrests tied to AI-generated abuse material. Law enforcement bodies say the spread of accessible AI tools is forcing rapid changes in forensic investigation and in the handling of digital evidence.

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

Google adds Personal Intelligence to AI Search

Google has expanded AI Search with Personal Intelligence, enabling more personalised responses using Gmail and Google Photos data. The feature aims to combine global information with individual context to deliver search results tailored to each user.

Eligible Google AI Pro and AI Ultra subscribers can opt in to securely connect their Gmail and Photos accounts, allowing Search to draw on personal preferences, travel plans, purchases, and memories.

The system uses contextual insights to generate recommendations that reflect users’ habits, interests, and upcoming activities.

Personal Intelligence enhances shopping, travel planning, and lifestyle discovery by anticipating needs and offering customised suggestions. Privacy controls remain central, with users able to manage data connections and turn off personal context at any time.

The feature is launching as an experimental Labs release for English-language users in the United States, with broader availability expected following testing. Google said ongoing feedback will guide refinements as the system continues to evolve.

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

Education for Countries programme signals OpenAI push into public education policy

OpenAI has launched the Education for Countries programme, a new global initiative designed to support governments in modernising education systems and preparing workforces for an AI-driven economy.

The programme responds to a widening gap between rapid advances in AI capabilities and people’s ability to use them effectively in everyday learning and work.

Education systems are positioned at the centre of closing that gap, as research suggests a significant share of core workplace skills will change by the end of the decade.

By integrating AI tools, training and research into schools and universities, national education frameworks can evolve alongside technological change and better equip students for future labour markets.

The programme combines access to tools such as ChatGPT Edu and advanced language models with large-scale research on learning outcomes, tailored national training schemes and internationally recognised certifications.

A global network of governments, universities and education leaders will also share best practices and shape responsible approaches to AI use in classrooms.

Initial partners include Estonia, Greece, Italy, Jordan, Kazakhstan, Slovakia, Trinidad and Tobago and the United Arab Emirates. Early national rollouts, particularly in Estonia, already involve tens of thousands of students and educators, with further countries expected to join later in 2026.

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