AI brings Babylon’s lost hymn back to life

A hymn to the ancient city of Babylon has been reconstructed after 2,100 years using AI to piece together 30 clay tablet fragments. Once lost after Alexander the Great’s conquest, the song praises the city’s grandeur, morals and daily life in exceptional poetic detail.

The hymn, sung to the god Marduk, depicts Babylon as a flourishing paradise filled with jewelled gates, verdant pastures and flowing rivers. AI tools helped researchers quickly assemble and translate the fragments, revealing a third of the original 250-line text.

The poem sheds rare light on Babylonian values, highlighting kindness to foreigners, the release of prisoners and the sanctity of orphans. It also gives a surprising glimpse into the role of women, including cloistered priestesses who acted as midwives.

Parts of the hymn were copied out by schoolchildren up to 1,400 years after it was composed, showing its cultural importance. Scholars now place it alongside the Epic of Gilgamesh as one of the most treasured literary works from ancient Mesopotamia.

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

TikTok struggles to stop the spread of hateful AI videos

Google’s Veo 3 video generator has enabled a new wave of racist AI content to spread across TikTok, despite both platforms having strict policies banning hate speech.

According to MediaMatters, several TikTok accounts have shared AI-generated videos promoting antisemitic and anti-Black stereotypes, many of which still circulated widely before being removed.

These short, highly realistic videos often included offensive depictions, and the visible ‘Veo’ watermark confirmed their origin from Google’s model.

While both TikTok and Google officially prohibit the creation and distribution of hateful material, enforcement has been patchy. TikTok claims to use both automated systems and human moderators, yet the overwhelming volume of uploads appears to have delayed action.

Although TikTok says it banned over half the accounts before MediaMatters’ findings were published, harmful videos still managed to reach large audiences.

Google also maintains a Prohibited Use Policy banning hate-driven content. However, Veo 3’s advanced realism and difficulty detecting coded prompts make it easier for users to bypass safeguards.

Testing by reporters suggests the model is more permissive than previous iterations, raising concerns about its ability to filter out offensive material before it is created.

With Google planning to integrate Veo 3 into YouTube Shorts, concerns are rising that harmful content may soon flood other platforms. TikTok and Google appear to lack the enforcement capacity to keep pace with the abuse of generative AI.

Despite strict rules on paper, both companies are struggling to prevent their technology from fuelling racist narratives at scale.

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

Artists explore meaning and memory at Antwerp Art Weekend

At Antwerp Art Weekend, two standout exhibitions by Eddie Peake and the Amsterdam-based collective Metahaven explored how meaning shifts or falls apart in an age shaped by AI, identity, and emotional complexity.

Metahaven’s film follows a character interacting with an AI assistant while exploring poetry by Eugene Ostashevsky. It contrasts AI’s predictive language models with the unpredictable nature of poetry, using visual metaphors to expose how AI mimics language without fully grasping it.

Meanwhile, Peake’s immersive installation at TICK TACK turned the Belgian gallery into a psychological labyrinth, combining architectural intrusion, raw paintings, and a haunting audio piece. His work considers the weight of identity, sexuality, and memory, moving from aggression to vulnerability.

Despite their differences, both projects provoke questions about how language, identity, and emotion are formed and fractured. Each invites viewers to reconsider the boundaries of expression in a world increasingly influenced by AI and abstraction.

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

Meta’s AI chatbots are designed to initiate conversations and enhance user engagement

Meta is training AI-powered chatbots that can remember previous conversations, send personalised follow-up messages, and actively re-engage users without needing a prompt.

Internal documents show that the company aims to keep users interacting longer across platforms like Instagram and Facebook by making bots more proactive and human-like.

Under the project code-named ‘Omni’, contractors from the firm Alignerr are helping train these AI agents using detailed personality profiles and memory-based conversations.

These bots are developed through Meta’s AI Studio — a no-code platform launched in 2024 that lets users build customised digital personas, from chefs and designers to fictional characters. Only after a user initiates a conversation can a bot send one follow-up, and that too within a 14-day window.

Bots must match their assigned personality and reference earlier interactions, offering relevant and light-hearted responses while avoiding emotionally charged or sensitive topics unless the user brings them up. Meta says the feature is being tested and rolled out gradually.

The company hopes it will not only improve user retention but also serve as a response to what CEO Mark Zuckerberg calls the ‘loneliness epidemic’.

With revenue from generative AI tools projected to reach up to $3 billion in 2025, Meta’s focus on more prolonged and engaging chatbot interactions appears to be as strategic as social.

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

X to test AI-generated Community Notes

X, the social platform formerly known as Twitter, is preparing to test a new feature allowing AI chatbots to generate Community Notes.

These notes, a user-driven fact-checking system expanded under Elon Musk, are meant to provide context on misleading or ambiguous posts, such as AI-generated videos or political claims.

The pilot will enable AI systems like Grok or third-party large language models to submit notes via API. Each AI-generated comment will be treated the same as a human-written one, undergoing the same vetting process to ensure reliability.

However, concerns remain about AI’s tendency to hallucinate, where it may generate inaccurate or fabricated information instead of grounded fact-checks.

A recent research paper by the X Community Notes team suggests that AI and humans should collaborate, with people offering reinforcement learning feedback and acting as the final layer of review. The aim is to help users think more critically, not replace human judgment with machine output.

Still, risks persist. Over-reliance on AI, particularly models prone to excessive helpfulness rather than accuracy, could lead to incorrect notes slipping through.

There are also fears that human raters could become overwhelmed by a flood of AI submissions, reducing the overall quality of the system. X intends to trial the system over the coming weeks before any wider rollout.

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

Cloudflare’s new tool lets publishers charge AI crawlers

Cloudflare, which powers 20% of the web, has launched a new marketplace called Pay per Crawl, aiming to redefine how website owners interact with AI companies.

The platform allows publishers to set a price for AI crawlers to access their content instead of allowing unrestricted scraping or blocking. Website owners can decide to charge a micropayment for each crawl, permit free access, or block crawlers altogether, gaining more control over their material.

Over the past year, Cloudflare introduced tools for publishers to monitor and block AI crawlers, laying the groundwork for the marketplace. Major publishers like Conde Nast, TIME and The Associated Press have joined Cloudflare in blocking AI crawlers by default, supporting a permission-based approach.

The company also now blocks AI bots by default on all new sites, requiring site owners to grant access.

Cloudflare’s data reveals that AI crawlers scrape websites far more aggressively than traditional search engines, often without sending equivalent referral traffic. For example, OpenAI’s crawler scraped sites 1,700 times for every referral, compared to Google’s 14 times.

As AI agents evolve to gather and deliver information directly, it raises challenges for publishers who rely on site visits for revenue.

Pay per Crawl could offer a new business model for publishers in an AI-driven world. Cloudflare envisions a future where AI agents operate with a budget to access quality content programmatically, helping users synthesise information from trusted sources.

For now, both publishers and AI companies need Cloudflare accounts to set crawl rates, with Cloudflare managing payments. The company is also exploring stablecoins as a possible payment method in the future.

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

Qantas cyber attack sparks customer alert

Qantas is investigating a major data breach that may have exposed the personal details of up to six million customers.

The breach affected a third-party platform used by the airline’s contact centre to store sensitive data, including names, phone numbers, email addresses, dates of birth and frequent flyer numbers.

The airline discovered unusual activity on 30 June and responded by immediately isolating the affected system. While the full scope of the breach is still being assessed, Qantas expects the volume of stolen data to be significant.

However, it confirmed that no passwords, PINs, credit card details or passport numbers were stored on the compromised platform.

Qantas has informed the Australian Federal Police, the Cyber Security Centre and the Office of the Information Commissioner. CEO Vanessa Hudson apologised to customers and urged anyone concerned to call a dedicated support line. She added that airline operations and safety remain unaffected.

The incident follows recent cyber attacks on Hawaiian Airlines, WestJet and major UK retailers, reportedly linked to a group known as Scattered Spider. The breach adds to a growing list of Australian organisations targeted in 2025, in what privacy authorities describe as a worsening trend.

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

UK urged to prepare for agentic AI in government

Agentic AI, a new generation of AI that goes beyond automation to deliver full task orchestration, could change how government operates. Sharon Moore, CTO Public Sector UK at IBM, argues the UK Government must adopt this technology to drive operational efficiency and better public services.

Departments using AI agents have already recorded significant savings, such as 3,300 hours saved in HR tasks by East and North Hertfordshire NHS Trust and 800 hours monthly by a New Jersey agency. IBM itself has cut development costs by billions, showcasing the potential for large-scale productivity gains.

Agentic systems integrate multiple AI models and tools, solving complex problems with minimal human intervention. Unlike traditional chatbots, these systems handle end-to-end tasks and adapt across use cases, from citizen services to legacy software modernisation.

To implement these systems safely, the UK must address risks like data leaks, hallucinations, and compliance failures. Moore emphasises that future governance must shift from overseeing individual models to managing entire AI systems, built on transparency, security, and performance oversight.

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

Tinder trials face scans to verify profiles

Tinder is trialling a facial recognition feature to boost user security and crack down on fraudulent profiles. The pilot is currently underway in the US, after initial launches in Colombia and Canada.

New users are now required to take a short video selfie during sign-up, which will be matched against profile photos to confirm authenticity. The app also compares the scan with other accounts to catch duplicates and impersonations.

Verified users receive a profile badge, and Tinder stores a non-reversible encrypted face map to aid in detection. The company claims all facial data is deleted when accounts are removed.

The update follows a sharp rise in catfishing and romance scams, with over 64,000 cases reported in the US last year alone. Other measures introduced in recent years include photo verification, ID checks and location-sharing tools.

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

The cognitive cost of AI: Balancing assistance and awareness

The double-edged sword of AI assistance

The rapid integration of AI tools like ChatGPT into daily life has transformed how we write, think, and communicate. AI has become a ubiquitous companion, helping students with essays and professionals streamline emails.

However, a new study by MIT raises a crucial red flag: excessive reliance on AI may come at the cost of our own mental sharpness. Researchers discovered that frequent ChatGPT users showed significantly lower brain activity, particularly in areas tied to critical thinking and creativity.

The study introduces a concept dubbed ‘cognitive debt,’ a reminder that while AI offers convenience, it may undermine our cognitive resilience if not used responsibly.

MIT’s method: How the study was conducted

The MIT Media Lab study involved 54 participants split into three groups: one used ChatGPT, another used traditional search engines, and the third completed tasks unaided. Participants were assigned writing exercises over multiple sessions while their brain activity was tracked using electroencephalography (EEG).

That method allowed scientists to measure changes in alpha and beta waves, indicators of mental effort. The findings revealed a striking pattern: those who depended on ChatGPT demonstrated the lowest brain activity, especially in the frontal cortex, where high-level reasoning and creativity originate.

Diminished mental engagement and memory recall

One of the most alarming outcomes of the study was the cognitive disengagement observed in AI users. Not only did they show reduced brainwave activity, but they also struggled with short-term memory.

Many could not recall what they had written just minutes earlier because the AI had done most of the cognitive heavy lifting. This detachment from the creative process meant that users were no longer actively constructing ideas or arguments but passively accepting the machine-generated output.

The result? A diminished sense of authorship and ownership over one’s own work.

Homogenised output: The erosion of creativity

The study also noted a tendency for AI-generated content to appear more uniform and less original. While ChatGPT can produce grammatically sound and coherent text, it often lacks the personal flair, nuance, and originality that come from genuine human expression.

Essays written with AI assistance were found to be more homogenised, lacking distinct voice and perspective. This raises concerns, especially in academic and creative fields, where originality and critical thinking are fundamental.

The overuse of AI could subtly condition users to accept ‘good enough’ content, weakening their creative instincts over time.

The concept of cognitive debt

‘Cognitive debt’ refers to the mental atrophy that can result from outsourcing too much thinking to AI. Like financial debt, this form of cognitive laziness builds over time and eventually demands repayment, often in the form of diminished skills when the tool is no longer available.

Typing

Participants who became accustomed to using AI found it more challenging to write without it later on. The reliance suggests that continuous use without active mental engagement can erode our capacity to think deeply, form complex arguments, and solve problems independently.

A glimmer of hope: Responsible AI use

Despite these findings, the study offers hope. Participants who started tasks without AI and only later integrated it showed significantly better cognitive performance.

That implies that when AI is used as a complementary tool rather than a replacement, it can support learning and enhance productivity. By encouraging users to first engage with the problem and then use AI to refine or expand their ideas, we can strike a healthy balance between efficiency and mental effort.

Rather than abstinence, responsible usage is the key to retaining our cognitive edge.

Use it or lose it

The MIT study underscores a critical reality of our AI-driven era: while tools like ChatGPT can boost productivity, they must not become a substitute for thinking itself. Overreliance risks weakening the faculties defining human intelligence—creativity, reasoning, and memory.

The challenge in the future is to embrace AI mindfully, ensuring that we remain active participants in the cognitive process. If we treat AI as a partner rather than a crutch, we can unlock its full potential without sacrificing our own.

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