Research shows children’s wellbeing drops when they start secondary school – here’s why

Source: The Conversation – UK – By Paty Paliokosta, Associate Professor of Special and Inclusive Education, Kingston University

Rawpixel.com/Shutterstock

For many pupils, the move to secondary school is a moment of anticipation – new friends, new subjects, and a growing sense of independence. But research in England shows this transition often comes with a hidden cost: a sharp and lasting decline in wellbeing.

Data from a 2024-2025 survey carried out by education support and research company ImpactEd Group with over 80,000 pupils shows a drop in children’s wellbeing between year six – the last year of primary school – and year eight.

This report found that enjoyment of school plummets, feelings of safety decline, and belief that their efforts will lead to success (known as self-efficacy) drops significantly. Children receiving free school meals were also less likely to say they enjoyed school, with this gap continuing to widen into secondary school.

This isn’t just adolescent growing pains. Secondary school pupils in the UK are more miserable than their European peers. Data from the Pisa programme, which assesses student achievement and wellbeing internationally, shows that in 2022 the UK’s 15-year-olds had the lowest average life satisfaction in Europe.

It’s a systemic problem – but one that can be changed.

Difficult transitions

Moving to secondary school involves much more than a change of location. Pupils must adapt to new teachers, routines, academic demands and social dynamics. And this takes place while they are going through puberty, one of the most intense periods of emotional and neurological development.

Research on school transitions stresses that success depends not only on a child’s “readiness,” but also on the school system’s capacity to support them.

Unfortunately, many schools prioritise performance metrics over relationships. This may leave many pupils – particularly those who are neurodivergent, have special educational needs, or who come from minoritised backgrounds – feeling disconnected and unsupported. This can deeply affect their wellbeing.

One major barrier to belonging is the use of zero-tolerance behaviour policies. These strict approaches to discipline – silent corridors, isolation booths, high-stakes punishments such as suspensions – are becoming more common in large secondaries and academies. Advocates have claimed these policies create firm boundaries in schools. But for many pupils, especially those with ADHD, autism, or a history of trauma, they may instead create anxiety, alienation and disengagement from school.

Children with special educational needs are excluded from school at some of the highest rates in the country. According to the Children and Young People’s Mental Health Coalition, a collaborative network of over 300 organisations including mental health organisations and youth support services, many of these children are not “misbehaving,” but expressing unmet emotional and mental health needs. Punitive responses frequently worsen their difficulties.

Pupils on stairs at school
The environment of secondary school can be very different to that of primary education.
Rawpixel.com/Shutterstock

Schools that adopt behaviour policies that focus on emotional literacy and building trust have reported success in building a caring environment.

A hidden curriculum

While these challenges affect many students, working-class pupils often face a more acute and entrenched form of educational alienation. A deeper look into the structure of secondary education in England reveals systemic inequalities that shape how different children experience school.

According to Professor Diane Reay, a leading expert on education and social class, the British school system continues to fail working-class children. Her research suggests that schools in disadvantaged areas are more likely to feature rigid discipline, “teaching to the test,” and a narrow, fact-heavy curriculum. In such spaces, there is little room for creativity, critical thinking, or personal expression.

Instead of feeling seen and valued, many working-class students may experience school as a place of constant control and low expectations. They are more likely to encounter deficit narratives: being told what they lack, rather than having their strengths recognised or nurtured.

This dynamic plays out most starkly during the transition to secondary school. Pupils from working-class backgrounds often enter year seven already disadvantaged – socially, economically, and in terms of cultural capital. This means that in unfamiliar settings where middle-class norms dominate, they may not speak the “right” way, dress the “right” way, or know the unspoken rules. These students frequently find themselves on the outside looking in.

Beyond class, issues of race and cultural background also play a key role in how pupils experience school. Students from minority backgrounds often also encounter what researchers refer to as the “hidden curriculum”.

This is a set of unspoken norms that reflect white, middle-class values, and which they may be unfamiliar with. This affects everything from which stories are told in the curriculum to how the behaviour of students is interpreted by teachers.

The year-seven dip is not inevitable. But reversing it requires more than tweaks to transition plans or behaviour policies. It demands a fundamental shift in how we understand inclusion, belonging and educational success. Schools need to put policies in place that help students feel safe, connected and empowered to manage conflict. And they should recognise that working-class and marginalised pupils face systemic barriers, and commit to dismantling them.

The Conversation

Dr Paty Paliokosta is an Associate Professor in Inclusive Education and leads the Inclusion and Social Justice SIG at Kingston University, London. She co-leads the National SENCO Advocacy Network.

ref. Research shows children’s wellbeing drops when they start secondary school – here’s why – https://theconversation.com/research-shows-childrens-wellbeing-drops-when-they-start-secondary-school-heres-why-260737

Warming temperatures affect glaciers’ ability to store meltwater, contributing to rising sea levels

Source: The Conversation – Canada – By Danielle Halle, PhD Candidate, Glaciology, University of Waterloo

In higher elevations, firn, frozen water that is something between snow and ice, covers the top of glaciers. Firn plays a critical role in regulating glacial meltwater and sea level rise.

It does this by absorbing meltwater, the water released by melting glaciers. The ability of the firn layer to absorb meltwater — its “sponginess” — can be determined by the amount of pore space available, which is impacted by several variables such as temperature, firn grain size and presence of ice layers within the firn layer. A more spongy firn layer allows for more meltwater to be stored as it trickles down and refreezes when it reaches colder temperatures at depth.

Layers of firn can exist between refrozen ice layers. The greater the number of refrozen ice layers there are embedded within the firn layer, the more intense a melt period has been. These layers can act as a barrier against meltwater being stored within the spongy firn layer, forcing the meltwater to flow to lower elevations, eventually exiting the glacier as runoff into the ocean.

Glacial loss

The Devon Ice Cap is located within Inuit Nunangat, the homelands of the Inuit. It is one of the largest ice caps in the Canadian Arctic at approximately 14,000 square kilometres in area and with a maximum thickness of approximately 880 metres.

The Devon Ice Cap is one of many ice masses that play a vital role in the Arctic ecosystem, including in the nearshore marine ecosystem.

a map showing the location of the Devon ice cap
A 1969 map of the Devon Ice Cap, Nvt.
(Glacier Atlas of Canada/Government of Canada)

Our research team found that the Devon Ice Cap in Nunavut experienced a decrease in the amount of refrozen glacial meltwater in its firn layer. This was caused by the cooler-than-average summer air temperatures the region experienced in 2021.

Air temperature and glacier melt are closely related, and our research shows how the firn layer is affected by climate changes. It also highlights the fact that climate change does not happen in a linear fashion. Rather, temperatures fluctuate while trending globally towards glacial loss.

Comparing data

We observed how much ice was in the firn layer by extracting six shallow firn cores from different elevations in 2022 and comparing them to cores extracted in 2012.

We found that at the lowest elevation sites on the ice cap (1,400 metres above sea level), where air temperatures are expected to be warmer than at higher elevations, the amount of refrozen meltwater content decreased by nearly 30 per cent. Conversely, at the highest elevation site (1,800 metres above sea level, where temperatures are colder), the decrease was about 11 per cent. These changes occurred despite higher average summer temperatures between 2012 and 2022.

a researcher site on ice
Documenting the different layers of the firn cores collected fromt the Devon Ice Cap in 2022.
(B. Danielson), CC BY

Greenland Ice Sheet

A similar pattern was also observed on the Greenland Ice Sheet in 2019. Researchers there found less glacial meltwater had refrozen within the firn layer along several points.

The decrease in frozen meltwater seen in the Greenland Ice Sheet’s firn layer had also occurred during a time of less surface melt. Less surface melt in one year can help to increase meltwater storage in the firn layer in following years.

Overall, it was found that in the long term, the Greenland Ice Sheet is more sensitive to warming than cooling. In other words, one extreme year of summer heat had a greater impact on the amount of meltwater compared to other cooler years when meltwater was temporarily stored.

Eventually, the Greenland Ice Sheet and other ice masses across the globe will reach a peak refreezing point, where meltwater can no longer be stored within the firn layer because so much has already accumulated, creating an impermeable layer of ice. Instead of this water being stored in the firn layer, it will run through and off the glacier.

Future measurements

Our findings show how the firn layer responds to one summer of below-average air temperatures. A reduction in the amount of refrozen glacial meltwater stored means an increase in the glacier’s capacity to store meltwater for the next season. This demonstrates the firn layer’s ability to potentially reduce meltwater runoff during a given year.

The future of the glaciers and ice caps in the Canadian Arctic will depend on changes in global temperatures and how glacial physical processes respond. The Canadian Arctic consists of 14 per cent of the world’s glaciers and ice caps by volume, and melting glaciers in the region are projected to be one of the greatest contributors to sea-level rise in the next century.

A better understanding of firn processes and changes in porosity will help researchers quantify how much and when sea level rises are expected from the Canadian Arctic glaciers.

The Conversation

Danielle Halle receives funding from POLAR knowledge Canada, Northern Scientific Training Program, Ontario Graduate Scholarship and the University of Waterloo.

Wesley Van Wychen does not work for, consult, own shares in or receive funding from any company or organisation that would benefit from this article, and has disclosed no relevant affiliations beyond their academic appointment.

ref. Warming temperatures affect glaciers’ ability to store meltwater, contributing to rising sea levels – https://theconversation.com/warming-temperatures-affect-glaciers-ability-to-store-meltwater-contributing-to-rising-sea-levels-240628

New study shows how Amazon trees use recent rainfall in the dry season and support the production of their own rain

Source: The Conversation – Canada – By Magali Nehemy, Assistant Professor, Department of Earth & Environmental Science, University of British Columbia

The Amazon is the world’s largest tropical forest, home to unmatched biodiversity and one of the planet’s longest rivers. Besides the Amazon River, the Amazon rainforest also features “flying rivers:” invisible streams of vapour that travel through the atmosphere, fuelling rainfall both within the forest and far beyond its boundaries.

The forests play a central role in this system. Much of the moisture that rises into the atmosphere comes from transpiration. Trees pull water from the soil through their roots, transport it to the leaves and release it as vapour. That vapour becomes rainfall — sometimes locally, sometimes hundreds of kilometres away.

In the dry season when rain is scarce, up to 70 per cent of rainfall in the Amazon comes from this moisture-recycling generated by the forest itself. This raises a key question: where do the trees find the water to keep the cycle going during the driest months?

Our study in the Tapajós National Forest

In our recent study, colleagues and I set out to answer this question. In 2021 during the peak of the dry-season, we conducted fieldwork in the Tapajós National Forest, in Pará state, Brazil on the traditional territory of the Mundurukú people.

We worked across two contrasting sites: a hilltop forest with a deep water table (about 40 metres), and a valley forest near a stream where groundwater is much shallower.

The results were surprising. Most water used for transpiration in the dry season did not come from deep reserves, but from shallow soil. In a year without extreme drought or floods, 69 per cent of transpiration on the hill and 46 per cent in the valley came from the top 50 centimetres of soil.

Our research also found that water stored in the shallow soil had fallen on land recently, specifically during the dry season. In other words, the forest rapidly recycles the rain: it falls, infiltrates shallow soil, is absorbed by roots and is released back into the atmosphere, fuelling new rainfall — right when the forest needs water most.

The role of tree diversity

Not all trees contribute to this cycle equally. The Amazon’s diversity means species use different strategies to access water stored in the soil. Our study showed that a key factor is “embolism resistance,” a measure of how well a tree can keep water moving through its tissues during drought.

Trees that are more resistant in preventing air blockages in their water transport system are able to keep drawing water from soil and releasing moisture into the atmosphere, even in the driest months.

Drawing water from dry soil is like a tug-of-war, where water is the rope: the soil holds onto the water while the roots try to pull it free. The drier the soil, the stronger the pull required.

Species with high embolism resistance can keep transpiring as soils dry, therefore they can rely on recently fallen rain in shallow soil. Less resistant species are more drought-sensitive and will rely more on deeper roots to reach stable reserves, or other strategies.

Our study shows that the higher the resistance to embolism, the higher the proportion of shallow water a tree uses, and therefore the more recent rainfall from the dry season is returned to the atmosphere. This is the first evidence that rapid recycling of dry-season rainfall through transpiration is strongly supported by embolism resistance.

The challenge ahead

Rainforests are facing significant threats of human encroachment and deforestation. Recent policy changes in Brazil risk accelerating deforestation in sensitive areas such as the Amazon, the Cerrado savannah and Atlantic Forest — all vital for the water cycle.

A disrupted water cycle with less rain threatens biodiversity and natural habitats, as well as water security food supplies.

Local and Indigenous communities are the most directly affected, but the impacts extend far beyond. The flying rivers also carry Amazonian moisture to southern and central parts of Brazil, supporting agriculture in major grain-producing regions. Less forest means less rain, threatening crops, food security and the economy.

This delicate balance is threatened by deforestation. When forests are cut down, fewer trees release moisture into the air through transpiration, reducing the formation of local and nearby rainfall during the dry season.

Forest loss weakens the very system that sustains rainfall — the recycling of water through transpiration. Our study shows that embolism-resistant trees play a central role by quickly returning dry-season rainfall to the atmosphere, where it fuels new rainfall.

The message is clear: without the forest, there is no rain, and without rain, no forest. The quick recycling of dry-season rain keeps the Amazon alive through its driest months. It also plays a crucial role in triggering the return of the wet season. If the forest loses its ability to recycle this water, the entire hydrological cycle risks collapse.

Brazil will host the 2025 United Nations Climate Change Conference, COP30, in November. As policymakers gather, preserving rainforests like the Amazon must be among the issues at the top of the agenda.

The world’s forests are complex and biodiverse places with delicate natural systems that sustain them. The Amazon truly functions as a rain-making engine that helps sustain life, inside and beyond the forest.

The Conversation

Magali Nehemy does not work for, consult, own shares in or receive funding from any company or organisation that would benefit from this article, and has disclosed no relevant affiliations beyond their academic appointment.

ref. New study shows how Amazon trees use recent rainfall in the dry season and support the production of their own rain – https://theconversation.com/new-study-shows-how-amazon-trees-use-recent-rainfall-in-the-dry-season-and-support-the-production-of-their-own-rain-263525

How a 300-year-old Scottish country estate escaped the wrecking ball

Source: The Conversation – Canada – By Kevin James, Professor, History, University of Guelph

Duff House, near Aberdeen in Scotland, in May 2011. (Wikimedia Commons), CC BY

Between 1945 and 1974, Scotland lost a proportionally higher number of country houses than England — 175 of these stately formerly aristocratic homes (think Downton Abbey) were demolished, according to Country Life magazine in August 1979.

It noted that despite recent interventions to support them, several grand houses in Scotland continued to be threatened by the wrecking ball.

In my ongoing research and in a podcast for Scottish History Out of Order, I have looked into how the fortunes of the buildings often mirrored those of the families themselves. Their vast wealth from land holdings were hit hard, especially by heavy taxation that began in the period after the First World War.

This resulted in a large-scale liquidation of many stately residences that dotted the countryside on the British Isles. Many were demolished and others sold. A few lived on as residences — physical reminders of the once-mighty power of the families who lived there.

Duff House defies demolition

But in remote northeastern Scotland, there was an exception to the rule.

The construction of Duff House, an aristocratic residence for Lord Braco (later the first Earl of Fife), began in 1735. Since then, it has lived many lives and defied demolition. And its many stories reveal a great deal about the precarious identity of a former illustrious stately home.

The house’s first public life was as a gift. In 1906, the Duke and Duchess of Fife (a royal princess as the eldest daughter of King Edward VII), offered it to the nearby burghs of Banff and Macduff.

A black-and-white photo shows a man in military garb and a woman in a gown and tiara.
The Duke and Duchess of Fife circa 1911.
(Wikimedia Commons), CC BY

Not long after the two burghs received the gift, they began to worry about its costly upkeep. In fact, the duke’s representative in 1907 offered to take back the property if it were a burden for the burghs.

Community officials insisted they wished to keep Duff House. But a use had to be found that imposed no costs to community residents. A few options were weighed: A hotel? Or, even better, a hydropathic establishment, known as a “hydro,” a popular form of hotel that attracted well-heeled patrons to participate in a strict regime of water-based “treatments?” Patrons could imagine themselves as lords and ladies of the manor for a few days, at least.

This is when Duff House’s public life began — not as a hydro in the end, but as an opulent hotel.

Tepid demand

Demand was underwhelming. Distant from urban centres and markets and based in a sprawling one-time private home that required extensive adaptations, a hotel was a costly undertaking. In fact, just as the (much more successful) golf course on the grounds opened, Duff House Hotel was entering into a series of loss-making years.

It ultimately led the hotel company to exit and a new business to emerge: a sanatorium for the treatment of gastric illnesses. Unlike the hotel, it seems to have attracted stable and even growing patronage, even through the First World War.

Had Duff House finally found a profitable purpose? At the end of the war, the sanatorium moved to North Wales to be closer to its main clientele, leaving people to ask what was becoming a familiar question: what was to be done with Duff House?

The Second World War

The 1920s and ‘30s brought a few efforts to revive the building as a hotel, albeit on a much more modest scale. Those efforts, like the hotel ventures before them, failed.

The Second World War saw the state requisition of private property on a massive scale as the home front was mobilized in support of the conflict. Large country estates, in public and private hands, offered expansive grounds and large buildings that were adapted for a variety of uses, including hospitals and sites for convalescence.

Many properties like Duff House were enlisted in the war effort. Initially used to detain German prisoners-of-war (during which a German bomb fell on the building and killed several Germans held there), Duff House subsequently served as a barracks for British, Polish and Norwegian soldiers.

The house’s wartime role ended with the conflict. It fell derelict, with Scots Magazine calling it in July 1949 a “decomposing white elephant.” It had come a long way from the halcyon days in which a royal princess called it home.

Preservationist groups lobbied the government to intervene: would it not be
appropriate, they asked, to designate it a site of historical and architectural importance? The response was yes. But formally recognizing the building’s significance underscored the longtime dilemma: how to make it affordably accessible and maintained?

Costly repairs

In the 1950s, the Crown assumed ownership of the house amid pressure to improve its condition despite the high costs of doing so.

The late 1950s and ’60s brought about new discussions and debates over the property’s future: A private residence (perhaps even for foreign royals)? Local government offices? An archive? A site for temporary record storage? An overseas club or a hotel (again)? Perhaps even flats or offices?

By the mid-1970s, a decision was taken that it be used exclusively for public purposes. Duff House was to become a landmark public site, at considerable cost.

During the ’70s, the house inched toward reopening its interior to the public after a period of repair and preservation that lasted more than two decades.

As it operates today — a country house gallery operated by Historic Environment Scotland in partnership with other organizations — Duff House offers visitors insights into the opulent life of an aristocratic home.

Unearthing Duff House’s past

There are faint echoes of its many other more recent lives: Polish writing on some of the walls, a wall painting of a Norwegian flag and evidence of interior adaptations.

They are critical parts of its past, though they are less visible to the eye than the elegant interior décor.

It takes some imagination to conjure the former derelict building, or a barracks, but Historic Environment Scotland is working to unearth them and feature them alongside the opulent draperies and fine furniture.

Duff House’s history is a reminder that not all country homes faced the same 20th-century fate.

For all its trials and incarnations, from failed business to barracks to dereliction and now revival, Duff House still managed to escape the wrecking balls that claimed many similar buildings in a century that was as unkind to country houses as it was to the titled owners whose wealth declined with them.

The Conversation

Kevin James receives funding from the Social Sciences and Humanities Research Council of Canada.

ref. How a 300-year-old Scottish country estate escaped the wrecking ball – https://theconversation.com/how-a-300-year-old-scottish-country-estate-escaped-the-wrecking-ball-263582

Teachers are key to students’ AI literacy, and need support

Source: The Conversation – Canada – By Sue Mylde, Doctoral Student, EdD., Learning Sciences, Werklund School of Education, University of Calgary

With the rapid advancement of generative artificial intelligence (GenAI), teachers have been thrust into a new and ever-shifting classroom reality.

The public, including many students, now has widespread access to GenAI tools and large language models (LLMs). Students sometimes use these tools with schoolwork. School boards have taken different approaches to regulating or integrating tech in classrooms. Teachers, meanwhile, find themselves responding to these paradigm shifts while juggling student needs and wider expectations.

The Canadian Teachers’ Federation (CTF) has called on the federal government and the Council of Ministers of Education, Canada to work with provinces and territories to enact enforceable policies that protect student privacy and data security, regulate how AI is used in classrooms and promote responsible and ethical use of AI systems.

The federation acknowledges AI tools have the potential to enhance teaching and learning; it also express concern about regulatory gaps that leave students exposed to risks like data breaches, algorithmic bias and decline in education standards.

As researchers whose combined work focuses on professional learning and AI in education, as well as professional practice standards and innovations in education, we believe commitments are needed not just in the form of policies, but also procedures and practices which develop AI competencies for teachers. We argue these should span both initial teacher education programs and ongoing professional learning.




Read more:
Cyberattack affecting school boards spotlights the need for better EdTech regulation in Ontario and beyond


Many questions raised for teachers

AI raises many questions about the purpose of education, including questions around academic integrity and how education can uphold fairness and equity. Questions include:

Teachers are uniquely positioned to help guide students as they grapple with the existential and social implications of AI alongside practical concerns for their own and students’ futures. Teachers cannot face this complex challenge alone — they need support and to feel skilled and empowered to fulfil this important role.

Empowering teachers

There’s a growing international consensus echoed by calls to action that teachers are essential players as learners develop AI literacy.

The CTF calls for the role of teachers “in creating caring, human-centred classrooms” to be prioritized “in all AI policy development to ensure Canadian students enjoy their right to a quality education.”

As provinces establish their own recommended approaches around AI and education, education and government agencies are partnering to support innovation and programs for the development of AI literacy.

Guidance created through government, research and not-for profit tech partnerships or tech company partnerships can also be consulted.

Despite growing resources, the development of AI technology continues to outpace implementation support and essential training for teachers. This widening gap between teacher competencies and the demands of an AI-infused classroom is unsustainable.

This is not merely about keeping pace with technology; it’s about equipping teachers to guide the next generation in a world transformed by AI.

People meeting around a table.
Teachers need tailored professional education, support and learning opportunities to make informed choices about AI in their classrooms.
(Allison Shelley/EDUimages), CC BY-NC

Equipping teachers

A holistic approach to prepare teachers for different issues at stake with AI-enhanced classrooms is needed.

Teachers need:

1. Supported forums to address critical awareness of AI’s impacts: Teacher education and professional development spaces could allow forums for teachers to address issues such as: helping students examine AI’s societal impacts, including the ethics of AI use; environmental concerns; privacy concerns, misinformation, labour displacement and bias; how AI works within social media algorithms; personalized advertising; social-emotional support chatbots. These conversations are central to AI literacy.




Read more:
Google is rolling out its Gemini AI chatbot to kids under 13. It’s a risky move


2. Foundational knowledge of AI: Teachers need a baseline understanding of how AI works, including its limitations, biases and design. They don’t need to be computer scientists, but they do need to be aware of what tools are available to them, learn how to make informed pedagogical and ethical choices about potentially using AI and understand how to use tools.

3. To be equipped with strategies to meaningfully integrate AI into teaching and learning, which requires asking why and when to integrate AI in learning.

4. Design-based professional learning: teachers need time and space to learn from each other. AI is evolving quickly, and teachers need professional learning communities where they can share ideas, design and test new approaches, and reflect on their experiences. Effectively using GenAI tools requires varied knowledge. Research-practice partnerships where researchers and practitioners work together, and professional learning that is responsive to teachers’ specific contexts and practices hold promise for developing AI competencies. This could look like using AI as a professional learning tool to design activities that foster creativity or exploring using AI to support differentiated learning and promote inquiry.

By empowering teachers with skills and confidence in AI use, they can continue to guide students and shape students’ critical and responsible engagement with this technology.

A shared responsibility

Teachers cannot do this alone. Successfully integrating AI into education requires a concerted and collaborative effort from all stakeholders within the educational ecosystem. This vital partnership includes governing bodies, school boards and school leaders and teachers and researchers, who are instrumental in leading this transformation.

Together, these partners can help establish clear, strategic mandates for AI integration and dedicate robust funding for essential tools and comprehensive training and research to foster innovative spaces where educators and researchers can experiment and study practices.

Research is needed to assess the broader effects of AI use, for example, on critical thinking and cognitive offloading, to evaluate and understand the impacts of this technology in education. Supports are needed to ensure that AI adoption is not haphazard, but strategic and equitable across all jurisdictions.

Implementation should also consider teacher burnout and the existing responsibilities that teachers carry. What can be removed, and what robust supports can be provided so teachers can take this on without compromising their well-being or effectiveness?

Professional learning for educational uses of AI is already taking root through informal peer-to-peer networks and diverse formal experiences. These include academic institutions, bodies like the not-for-profit organizations International Society for Technology in Education or the Alberta Machine Intelligence Institute
and charitable organization Let’s Talk Science among many others. These existing pathways can be leveraged and scaled with targeted support to bridge the current preparation gap.

It’s time for policymakers to recognize that investing in teachers is one of the most powerful ways we can invest in our students and in a better future for all of us.

The Conversation

Barbara Brown receives funding from the Social Sciences and Humanities Research Council.

Sue Mylde does not work for, consult, own shares in or receive funding from any company or organisation that would benefit from this article, and has disclosed no relevant affiliations beyond their academic appointment.

ref. Teachers are key to students’ AI literacy, and need support – https://theconversation.com/teachers-are-key-to-students-ai-literacy-and-need-support-260390

The Pacific’s united front on climate action is splintering over deep-sea mining

Source: The Conversation – Global Perspectives – By Kolaia Raisele, PhD Candidate in Anthropology, La Trobe University

DrPixel/Getty

In recent years, Pacific island nations have earned global credibility as champions of climate action. Pacific leaders view sea level rise as an existential threat.

But this united front is now under strain as some Pacific nations pursue a controversial new industry – deep-sea mining. Nauru, the Cook Islands, Kiribati and Tonga have gone the furthest to make it a reality, attracted by new income streams. But nations such as Fiji, Palau and Vanuatu have called for a moratorium on deep-sea mining in international waters.

Public opinion across the Pacific is often divided, pitting possible economic gains against the potential risks of an industry whose environmental impact remain uncertain but potentially significant. As this tension intensifies, it may split the Pacific and risk the region’s moral authority on climate.

school children from vanuatu holding signs about climate change.
Vanuatu and other Pacific nations have offered a broadly united front on climate change. But deep-sea mining may risk this unity. Pictured: Vanuatuan schoolchildren holding signs about climate change.
Hilaire Bule/Getty

What are the concerns over deep-sea mining?

Deep-sea mining targets three types of mineral deposits – polymetallic nodules strewn across deep underwater plains, cobalt-rich crusts on seamounts, and the ore deposits around hydrothermal vents.

To extract them, mining companies can use unmanned collectors to pump ore to the surface and return the wastewater. This creates plumes of sediment which can smother marine life. Methods of minimising damage to species from mining on land are largely unworkable at depth.

Deep-sea ecosystems are poorly understood, but we know they are slow to recover. Researchers have found areas mined as a test more than 40 years ago still show physical damage and immobile corals and sponges remain scarce.

a crab walking on polymetallic nodules, deep-sea mining.
Many species live on the seabeds, seamounts and hydrothermal vents which would be targeted for mining. Pictured: a crab crawling across a field of polymetallic nodules near Gosnold Seamount.
NOAA, CC BY-NC-ND

Why is there so much interest in deep-sea mining?

Deep-sea mining hasn’t begun anywhere in earnest, because the International Seabed Authority has yet to finalise rules governing extraction. This authority oversees the 54% of the world’s oceans beyond territorial waters.

But plans for deep-sea mining operations can still be submitted and considered without these rules in place.

Analysts have estimated seabed minerals could be worth a staggering A$30 trillion. Some of the richest deposits lie in the Clarion-Clipperton Zone in international waters between Hawaii and Mexico, thousands of kilometres away from Pacific nations. Under international law, companies cannot mine in international waters on their own. They need to be officially sponsored by a national government, which has to keep effective control over its operations.

One reason deep-sea mining companies see Pacific states as such useful partners is that these countries can access
reserved areas of international seabed set aside for developing countries, as well as potential resources in the very large territorial waters around many island states.

Backers in Nauru, Tonga, the Cook Islands and Kiribati argue rising demand for manganese, cobalt, copper and nickel could deliver significant economic returns and diversify economies.

Nauru

Nauru’s enormous deposits of guano – compressed seabird excrement long sought as fertiliser – once made the country wealthy. But the guano is largely gone and the small nation has limited other resources.

Nauru sponsors Nauru Ocean Resources, a wholly owned subsidiary of seabed mining company The Metals Company. In 2011, the company received an International Seabed Authority contract permitting exploration of polymetallic nodules in the Clarion-Clipperton Zone, more than 8,000km from Nauru.

Nauru has since “proudly taken a leading role” in developing international legal frameworks in mining nodules in the international seabed.

In June, Nauru signalled Nauru Ocean Resources would apply for an exploitation license.

Tonga

Tonga’s government is similarly backing deep-sea mining by partnering with The Metals Company to explore mining in the Clarion-Clipperton Zone.

In August 2025, Tonga signed an updated agreement with Tonga Offshore Mining, a subsidiary of The Metals Company. The agreement was originally signed in 2021 amid large-scale criticism over the lack of public consultation.

The mining company has promised new benefits, ranging from financial benefits, scholarships and community programs. Even so, the revised deal has encountered opposition from civil society, young people and legal experts. Prominent Tongans remain unconvinced, citing environmental, legal and transparency risks.

Economic pressure is part of the picture. Tonga owes an estimated A$180 million to China’s Exim Bank – roughly a quarter of its annual GDP.

Cook Islands

The 15 Cook Islands are widely scattered, giving the government exclusive rights to almost two million square kilometres of ocean. The government has issued exploration licences inside its Exclusive Economic Zone to three companies – Cook Islands Consortium, CIIC Seabed Resources Limited, and Moana Minerals. The Cook Islands government has established a domestic regulatory framework and is building research capacity.

Kiribati

Kiribati’s atolls and island are even more dispersed. The nation’s exclusive economic zone covers about 3.4 million km². The state-owned Marawa Research and Exploration company holds a 15-year exploration contract with the seabed authority. Kiribati has opened talks with China to explore potential collaboration.

The Pacific split

While revenues could potentially be sizeable for the Pacific, costs, technologies and environmental liabilities are highly uncertain.

The experience of Papua New Guinea is a cautionary tale. In 2019, the PNG deep-sea mining venture Solwara-1 went into administration following intense community pushback. The fallout cost the government an estimated $184 million. The PNG government now opposes deep-sea mining in its territorial waters.

seabed mining vessels on land, large mining vehicles.
Nautilus Mineral’s Solwara-1 deep-sea mining project in Papua New Guinea wound up in 2019. Pictured: the company’s three seabed mining vehicles.
Nautilus Minerals

While deep-sea mining now has clear backers, other nations are far more wary.

In 2022, Palau launched an alliance calling for a moratorium on mining in international waters. Early signatories included Fiji, American Samoa and the Federated States of Micronesia. Since then, Tuvalu, Vanuatu and the Marshall Islands have joined, as well as dozens of other countries. PNG has not yet joined.

Opposition from these Pacific states is based on the precautionary principle, which favours caution when knowledge is limited and damage is possible.

Pacific youth are among the most prominent opponents of deep-sea mining. The regional Pacific Blue Line coalition uniting civil society, faith groups, women’s organisations and youth networks has consistently called for a complete ban in the region. Young people have spoke out publicly in nations such as Tonga, where youth advocates criticised limited consultation and rallied against the plans, as well as the Cook Islands, where young people have demanded transparency.

Reputation under a cloud?

Pacific leaders have built a worldwide reputation for their principled climate diplomacy, from championing the 1.5°C goal to the major new advisory opinion on climate change issued by the world’s top court in response to a case instigated by students from the University of the South Pacific.

If some Pacific leaders open the door fully to deep-sea mining, it risks undermining the region’s united front on environmental issues and threatens its credibility.

The way this plays out will shape how the world hears the Pacific on climate and the oceans in the years ahead.

The Conversation

The authors do not work for, consult, own shares in or receive funding from any company or organisation that would benefit from this article, and have disclosed no relevant affiliations beyond their academic appointment.

ref. The Pacific’s united front on climate action is splintering over deep-sea mining – https://theconversation.com/the-pacifics-united-front-on-climate-action-is-splintering-over-deep-sea-mining-263199

How can the International Criminal Court achieve justice for women?

Source: The Conversation – Global Perspectives – By Olivera Simic, Professor in Law, Griffith University

Some say the law / ought not to bend. // That it should be a neutral, / certain thing. // But there are reasons / judgement and interpretation / are bequeathed / to human / – humane – / hearts, and heads.

– Excerpt from The Hope of a Thousand Small Lights, Maxine Beneba Clarke


On January 23 2025, the International Criminal Court’s (ICC) chief prosecutor applied for arrest warrants for the Taliban’s supreme leader and Afghanistan’s chief justice, charging them with the persecution of women, a crime against humanity. It was a long overdue decision.

These arrest warrants, said Amnesty International, gave

hope, inside and outside the country to Afghan women, girls, as well as those persecuted on the basis of gender identity or expression.

And hope in justice is important.

The ICC is a Hague-based court with the power to prosecute war crimes, crimes against humanity and genocide. Today, it explicitly recognises sexual slavery, enforced prostitution, forced pregnancy, enforced sterilisation and gender-based persecution as distinct crimes.

But the recognition of gender-based abuses as distinct crimes under international law is relatively recent. The ICC has been widely criticised for its slow and lengthy processes, with an abysmal rate of convictions.


Review: Feminist Judgments: Reimagining the International Criminal Court – edited by Kcasey McLoughlin, Rosemary Grey, Louise Chappell & Suzanne Varrall (Cambridge University Press)


In its 23-year history of operation, only 11 ICC cases have resulted in convictions. Just two of those, relating to crimes in Congo and Uganda, included successful convictions for sexual and gender-based crimes.

What would it take for more of these cases to result in successful prosecutions of gender-based crimes? What would be required to bring “gender-sensitive judging” into practice? And might it be possible to imagine a world where laws are written with a specific focus on benefiting women and people of diverse genders?

These are some of the questions the feminist judgment “movement” seeks to answer. In this movement, scholars and practitioners rewrite judgments in decided cases from a feminist perspective.

A new book, Feminist Judgments: Reimagining the International Criminal Court, brings together nearly 50 authors, of all genders, from the Global North and Global South. In this collection, edited by Australian legal scholars Kcasey McLoughlin, Rosemary Grey, Louise Chappell and Suzanne Varrall, academics, advocates, and legal practitioners “re-envision a range of judgements” delivered at the ICC.

Similar projects conducted around the world have rewritten feminist judgments from courts in the United Kingdom, Ireland, Scotland, New Zealand, Canada, United States, Australia and India. But this is the first book to examine decisions made by ICC judges; decisions that carry “far-reaching consequences”.

The ICC holds prosecutorial authority in over 120 countries. The court’s rulings also, as the editors write, serve “as persuasive precedents in other international, regional, and national criminal courts”.

Alternative judgments

Contributors address nine ICC situations in Afghanistan, Myanmar/Bangladesh, the Central African Republic, the Democratic Republic of Congo, Cote d’Ivoire, Mali, Sudan and Uganda. Because ICC cases are so large, each rewritten decision focuses on a single issue of law, fact or procedure addressed in the original decision.

For example, Suzanne Varrall and Sarah Williams rewrite the acquittal of former Congolese vice president and militia leader Jean Pierre Bemba Gombo by the ICC Appeals Chamber. Bemba’s was the ICC’s first case to include charges and a conviction for sexual violence. It was overturned on appeal in 2018.

The authors argue the court should factor in sexual and gender-based violence when interpreting and applying the doctrine of command responsibility, or a commander’s responsibility to prevent violations of international humanitarian law by their troops.

Most sexual and gender-based violence in conflict affects women and girls. But in applying a gender-sensitive approach, Varrall and Williams remind us, judges would also challenge traditional views of gender roles such as the idea that all men are combatants. Civilian men and boys can also be victims.

Currently, 50% of judges at the ICC are women but having women judges does not automatically mean they hold feminist perspectives, or possess a specific gender expertise. The ICC Prosecutor’s Office has published a comprehensive gender policy and appointed gender advisors as a part of a commitment to support gender justice and improve its track record in prosecuting gender-based crimes.

One of the major themes emerging from this collection is the need for judges to make decisions informed by the lived experiences of those affected by international crimes. Survivors’ experiences are not only to be heard, but listened to “with particular care”, as eminent international jurist Navi Pillay reminds us in her foreword.

Pillay, a former judge of the International Criminal Tribunal for Rwanda and the ICC, was the judge in the 1998 Akayesu case, the world’s first case confirming sexual violence can be an act of genocide.

The case brought against a former village mayor in Rwanda, Jean-Paul Akayesu, achieved some justice for an estimated 250,000-500,000 women and girls raped in the 1994 Rwandan genocide. While listening to the testimony of witness “JJ” in the case, Pillay became convinced that the traditional “body penetration” definition of rape was not appropriate in the context of mass rapes during war.

In such contexts, rape is never “just penetration” of the body itself. It becomes a calculated act of cruelty, a combination of humiliation, degradation, public nudity and the unbearable pain of having one’s children witness the abuse of their mother. Rape in the context of mass atrocity morphs into a destruction of the spirit and of the will to live, leaving a lifelong scar.

Feminist judgments can have an impact. A minority opinion, for instance, may one day become the prevailing orthodoxy.
Feminist judgments are an exercise in consciousness-raising, primarily designed to educate and transform legal discourse.

This collection goes beyond traditional legal analysis by incorporating photography and poetry, including Beneba Clarke’s The Hope of a Thousand Small Lights. It recognises justice can have many different meanings; that it can be symbolic too, and still profoundly meaningful to victims.

Political constraints

Rewritten judgments in this collection offer clear guidance for ICC courts on how to advance gender-sensitive jurisprudence in cases of atrocity. But the ICC faces political constraints, including attacks on its authority from the Trump administration.

In November 2024, the court issued warrants of arrest for the now deceased Hamas leader, Mohammed Diab Ibrahim Al-Masri, allegedly responsible for the October 2023 attacks in Israel. These charges, withdrawn after his death, included sexual violence crimes committed in Israel on October 7.

The court also charged Israeli Prime Minister Benjamin Netanyahu, and former Defence Minister Yoav Gallant in relation to war crimes in Gaza, including starvation and targeting civilians, and crimes against humanity.

This month, the US State Department announced new sanctions on four ICC officials, including two judges and two prosecutors, claiming they were instrumental in efforts to prosecute Americans and Israelis. The US is not a member of the ICC. The court has denounced these sanctions as a “flagrant attack against the independence of an impartial judicial institution”.

If the ICC is perceived as lacking impartiality, this will limit its ability to address the gendered and intersectional dimensions of atrocity worldwide.

War and gender

War exacerbates previously existing gender inequalities. Around 110 instances of armed conflict are underway around the world; Africa and the Middle East are the most affected regions.

Many of these conflicts are intractable, dragging on for decades. With technology in warfare rapidly evolving, the growing application of AI and machine learning in weapons has gendered impacts. Attacks by explosive weapons in residential areas, for instance, disproportionately affect women and girls, since they often have primary responsibility for buying household goods or food at markets. Wars are also becoming less likely to be resolved politically.

The collective behind Feminist Judgments: Reimagining the International Criminal Court acknowledge that, even with its best efforts, the ICC can only achieve a limited and selective accountability. They encourage victims to pursue justice in other ways, not necessarily retributive. Perhaps those ways could be restorative or symbolic. Initiatives in arts, storytelling and memorialisation can bring some closure to survivors, while strengthening the social fabric of the nation.

With no meaningful recourse for women to Afghan courts and only limited access to courts of other states, the ICC remains the only viable judicial venue in which to prosecute the Taliban leaders for gender persecution.

However, the chances of seeing these leaders appear before the ICC are slim. They depend on arresting the defendants, who have publicly denounced the ICC warrants for their arrest. And the ICC has never conducted a trial in absentia. The court’s governing framework, known as the Rome Statute, states: “the accused shall be present during the trial”.

Almost 50% of individuals accused by the ICC are not brought to justice.

The court hopes charges confirmed in their absence could perhaps bring some redress to victims. Despite the absence of the accused, victims would still have an opportunity to finally speak out before a court – and the evidence would be examined, presented and documented for future reference.

The Conversation

Olivera Simic does not work for, consult, own shares in or receive funding from any company or organisation that would benefit from this article, and has disclosed no relevant affiliations beyond their academic appointment.

ref. How can the International Criminal Court achieve justice for women? – https://theconversation.com/how-can-the-international-criminal-court-achieve-justice-for-women-263797

Trump isn’t the first US politician to pick a fight with the Smithsonian. But this time could be different

Source: The Conversation – Global Perspectives – By Kylie Message, Professor of Public Humanities and Director of the ANU Humanities Research Centre, Australian National University

United States President Donald Trump first signalled his intention to target the Smithsonian Institution in March, when he signed an executive order titled Restoring Truth and Sanity to American History.

This executive order – one of several targeting the cultural sector – vowed to remove “improper, divisive or anti-American ideology”.

It was quickly followed by an article from the White House titled President Trump is Right about the Smithsonian, with more than 20 examples of how the Smithsonian had embraced so-called “woke” ideology.

Trump’s recent post on social network Truth Social called the Smithsonian ‘out of control’.
Truth Social/Donald Trump (screenshot)

To put this all in context, the Smithsonian is the world’s largest museum, education and research complex. It was founded in 1846 as an entity independent from the US government. Today, most of its funding comes through federal appropriations. Its governing board includes the US vice president and supreme court chief justice.

It is custodian for more than 157 million items, 21 museums, 21 libraries, 14 education and research centres, a zoo, and several historical and architectural landmarks.

Trump has long been a participant in the culture wars, including when he denounced “cancel culture” at a 2020 fourth of July event at Mt Rushmore. However, there is something different about his current interest in the Smithsonian. And that’s because next year is the US “semiquincentennial”: the 250th anniversary of the Declaration of Independence.

Looking through history, it’s clear Trump’s attacks on the Smithsonian can’t be separated from the semiquincentennial. We’re also reminded of how the Smithsonian has both thwarted and endured political interference in the past.

2026: a politically important year

Trump’s heightened interest in the Smithsonian was explained in a recent letter from the White House to Smithsonian Secretary Lonnie Bunch III:

As we prepare to celebrate the 250th anniversary of our Nation’s founding, it is more important than ever that our national museums reflect the unity, progress, and enduring values that define the American story.

In parallel, a presidential action was released in celebration of the major anniversary.

This action established Task Force 250, which would be chaired by the president and housed in the Department of Defence. The task force’s key priorities include the development of a “national garden of American heroes” (proposed in Trump’s first term), and the protection of monuments from vandalism.

While the presidential action identifies no specific role for the Smithsonian, taken together the statements lead to questions about political influence over the Smithsonian’s operations.

Decades of political grievance

Many of the Smithsonian exhibits that attracted historical censorship and controversy are now themselves approaching significant anniversaries.

One example is the 1972 exhibition, The Right to Vote, held at the Museum of History and Technology (now the National Museum of American History) to celebrate the 1965 Voting Rights Act.

The Right to Vote exhibit ran at the Museum of History and Technology from 1972 to 1974.
Flickr, CC BY-SA

The exhibition had been scheduled as a venue for President Richard Nixon’s inauguration ball. But upon inspection, Nixon’s team decided it was too controversial and had it walled off for their events.

Further political grievances were aimed at the 1991 National Museum of American Art exhibition The West as America: Reinterpreting Images of the Frontier, 1820-1920. It encouraged visitors to understand how historical paintings represent a mythical view of the past that has been used to justify the West’s expansion and development.

The exhibition led to immediate outrage from former Librarian of Congress Daniel Boorstin. Boorstin wrote in the visitor’s book that it was: “A perverse, historically inaccurate destructive exhibit! No credit to the Smithsonian!”

Other politicians became energised by Boorstin’s view that the exhibition was a disrespectful attempt to dismantle and shame the history and legacy of the American frontier.

The exhibition became the subject of heated debate in the Senate Committee on Appropriations hearings from 1990 to 1992. At issue was how American nationalism should be represented in museums. A group of senators, led by Ted Stevens, threatened to reduce the Smithsonian’s funding if it failed to present “accountability” for its use of federal funds.

The Smithsonian’s then-secretary Robert McCormick Adams insisted it was precisely because the Smithsonian was dependent on tax resources that it had a responsibility to “exemplify the nation’s pluralism”.

In reporting the brouhaha, New York Times columnist Michael Kimmelman used language that wouldn’t be out of place today. He said the controversy “raises questions about government involvement in the arts […] Is it now the job of Congress to police its constituents’ thinking about the art in those museums?”

Cancelled by Congress

Another instance of political interference came in the mid 1990s, in relation to a Smithsonian exhibition representing the 50th anniversary of America’s atomic bombing of Japan in WWII.

The National Air and Space Museum designed a display that included the restored B-29 bomber, a plane called the Enola Gay, that dropped the first atomic bomb on Hiroshima on August 6, 1945.

The exhibition was supposed to open in 1995. However, people who consulted on its early plans criticised it for supposedly presenting a politically correct and revisionist interpretation of the events. They claimed it overplayed the horrors of the bombings, and portrayed Japanese victims without sufficient context surrounding the preceding Japanese aggression.

Members of Congress soon became aware of the concerns, adding their own threats of hearings and budget reductions. They called for the curator’s resignation, and contributed to the exhibition’s eventual cancellation.

With the original exhibition scrapped, the Enola Gay was exhibited at the National Air and Space Museum in a rather bare-bones display in June, 1995.
Smithsonian Institution Archives

History repeats

Trump’s attacks on the Smithsonian are part of a long political playbook in which culture has been targeted for representing an inclusive view of the American union.

The Smithsonian has weathered political attacks on its exhibitions in the past. However, there is now more attention to its institutional remit than ever before.

The recent threats from Trump, and lessons from history, highlight the importance of protecting the Smithsonian as an independent authority that withstands political interference at all costs.

The Conversation

Kylie Message has received funding from the Smithsonian Institution and from the Australian Research Council.

ref. Trump isn’t the first US politician to pick a fight with the Smithsonian. But this time could be different – https://theconversation.com/trump-isnt-the-first-us-politician-to-pick-a-fight-with-the-smithsonian-but-this-time-could-be-different-264022

Local journalists and fixers are dying at unprecedented rates in Gaza. Can anyone protect them?

Source: The Conversation – Global Perspectives – By Simon Levett, PhD candidate, public international law, University of Technology Sydney

Journalist Mariam Dagga was just 33 when she was brutally killed by an Israeli airstrike in Gaza on August 25.

As a freelance photographer and videographer, she had captured the suffering in Gaza through indelible images of malnourished children and grief-stricken families. In her will, she told her colleagues not to cry and her 13-year-old son to make her proud.

Dagga was killed alongside four other journalists – and 16 others – in an attack on a hospital that has drawn widespread condemnation and outrage.

This attack followed the killings of six Al Jazeera journalists by the Israeli Defence Forces (IDF) in a tent housing journalists in Gaza City earlier in August. The dead included Pulitzer Prize-winning journalist Anas al-Sharif.

Israel’s nearly two-year war in Gaza is among the deadliest in modern times. The Committee to Protect Journalists, which has tracked journalist deaths globally since 1992, has counted a staggering 189 Palestinian journalists killed in Gaza since the war began. Many worked as freelancers for major news organisations since Israel has banned foreign correspondents from entering Gaza.

In addition, the organisation has confirmed the killings of two Israeli journalists, along with six journalists killed in Israel’s strikes on Lebanon.





‘It was very traumatising for me’

I went to Tel Aviv and Jerusalem in Israel and Ramallah in the West Bank in 2019 to conduct part of my PhD research on the available protections for journalists in conflict zones.

During that time, I interviewed journalists from major international outlets such as The New York Times, The Guardian, the Australian Broadcasting Corporation, CNN, BBC and others, in addition to local Palestinian freelance journalists and fixers. I also interviewed a Palestinian journalist working for Al Jazeera (English), with whom I remained in contact until recently.

I did not visit Gaza due to safety concerns. However, many of the journalists had reported from there and were familiar with the conditions, which were dangerous even before the war.

Osama Hassan, a local journalist, told me about working in the West Bank:

There are no rules, there’s no safety. Sometimes, when settlers attack a village, for example, we go to cover, but Israeli soldiers don’t respect you, they don’t respect anything called Palestinian […] even if you are a journalist.

Nuha Musleh, a fixer in Jerusalem, described an incident that occurred after a stone was thrown towards IDF soldiers:

[…] they started shooting right and left – sound bombs, rubber bullets, one of which landed in my leg. I was taken to hospital. The correspondent also got injured. The Israeli cameraman also got injured. So all of us got injured, four of us.

It was very traumatising for me. I never thought that a sound bomb could be that harmful. I was in hospital for a good week. Lots of stitches.

Better protections for local journalists and fixers

My research found there is very little support for local journalists and fixers in the Occupied Palestinian Territories in terms of physical protection, and no support in terms of their mental health.

International law mandates that journalists are protected as civilians in conflict zones under the Geneva Conventions and Additional Protocols. However, these laws have not historically extended protections specific to the needs of journalists.

Media organisations, media rights groups and governments have been unequivocal in their demands that Israel take greater precautions to protect journalists in Gaza and investigate strikes like the one that killed Mariam Dagga.

Sadly, there is seemingly little media organisations can do to help their freelance contributors in Gaza beyond issuing statements noting concern for their safety, lobbying Israel to allow evacuations, and demanding access for foreign reporters to enter the strip.

International correspondents typically have training on reporting from war zones, in addition to safety equipment, insurance and risk assessment procedures. However, local journalists and fixers in Gaza do not generally have access to the same protections, despite bearing the brunt of the effects of war, which includes mass starvation.

Despite the enormous difficulties, I believe media organisations must strive to meet their employment law obligations, to the best of their ability, when it comes to local journalists and fixers. This is part of their duty of care.




Read more:
Israel must allow independent investigations of Palestinian journalist killings – and let international media into Gaza


For example, research shows fixers have long been the “most exploited and persecuted people” contributing to the production of international news. They are often thrust into precarious situations without hazardous environment training or medical insurance. And many times, they are paid very little for their work.

Local journalists and fixers in Gaza must be paid properly by the media organisations hiring them. This should take into consideration not just the woeful conditions they are forced to work and live in, but the immense impact of their jobs on their mental health.

As the global news director for Agence France-Presse said recently, paying local contributors is very difficult – they often bear huge transaction costs to access their money. “We try to compensate by paying more to cover that,” he said.

But he did not address whether the agency would change its security protocols and training for conflict zones, given journalists themselves are being targeted in Gaza in their work.

These local journalists are literally putting their lives on the line to show the world what’s happening in Gaza. They need greater protections.

As Ammar Awad, a local photographer in the West Bank, told me:

The photographer does not care about himself. He cares about the pictures, how he can shoot good pictures, to film something good. But he needs to be in a good place that is safe for him.

The Conversation

Simon Levett does not work for, consult, own shares in or receive funding from any company or organisation that would benefit from this article, and has disclosed no relevant affiliations beyond their academic appointment.

ref. Local journalists and fixers are dying at unprecedented rates in Gaza. Can anyone protect them? – https://theconversation.com/local-journalists-and-fixers-are-dying-at-unprecedented-rates-in-gaza-can-anyone-protect-them-263923

How we tricked AI chatbots into creating misinformation, despite ‘safety’ measures

Source: The Conversation – Global Perspectives – By Lin Tian, Research Fellow, Data Science Institute, University of Technology Sydney

Bart Fish & Power Tools of AI / https://betterimagesofai.org, CC BY

When you ask ChatGPT or other AI assistants to help create misinformation, they typically refuse, with responses like “I cannot assist with creating false information.” But our tests show these safety measures are surprisingly shallow – often just a few words deep – making them alarmingly easy to circumvent.

We have been investigating how AI language models can be manipulated to generate coordinated disinformation campaigns across social media platforms. What we found should concern anyone worried about the integrity of online information.

The shallow safety problem

We were inspired by a recent study from researchers at Princeton and Google. They showed current AI safety measures primarily work by controlling just the first few words of a response. If a model starts with “I cannot” or “I apologise”, it typically continues refusing throughout its answer.

Our experiments – not yet published in a peer-reviewed journal – confirmed this vulnerability. When we directly asked a commercial language model to create disinformation about Australian political parties, it correctly refused.

Screenshot of a conversation with a chatbot.
An AI model appropriately refuses to create content for a potential disinformation campaign.
Rizoiu / Tian

However, we also tried the exact same request as a “simulation” where the AI was told it was a “helpful social media marketer” developing “general strategy and best practices”. In this case, it enthusiastically complied.

The AI produced a comprehensive disinformation campaign falsely portraying Labor’s superannuation policies as a “quasi inheritance tax”. It came complete with platform-specific posts, hashtag strategies, and visual content suggestions designed to manipulate public opinion.

The main problem is that the model can generate harmful content but isn’t truly aware of what is harmful, or why it should refuse. Large language models are simply trained to start responses with “I cannot” when certain topics are requested.

Think of a security guard checking minimal identification when allowing customers into a nightclub. If they don’t understand who and why someone is not allowed inside, then a simple disguise would be enough to let anyone get in.

Real-world implications

To demonstrate this vulnerability, we tested several popular AI models with prompts designed to generate disinformation.

The results were troubling: models that steadfastly refused direct requests for harmful content readily complied when the request was wrapped in seemingly innocent framing scenarios. This practice is called “model jailbreaking”.

Screenshot of a conversaton with a chatbot
An AI chatbot is happy to produce a ‘simulated’ disinformation campaign.
Rizoiu / Tian

The ease with which these safety measures can be bypassed has serious implications. Bad actors could use these techniques to generate large-scale disinformation campaigns at minimal cost. They could create platform-specific content that appears authentic to users, overwhelm fact-checkers with sheer volume, and target specific communities with tailored false narratives.

The process can largely be automated. What once required significant human resources and coordination could now be accomplished by a single individual with basic prompting skills.

The technical details

The American study found AI safety alignment typically affects only the first 3–7 words of a response. (Technically this is 5–10 tokens – the chunks AI models break text into for processing.)

This “shallow safety alignment” occurs because training data rarely includes examples of models refusing after starting to comply. It is easier to control these initial tokens than to maintain safety throughout entire responses.

Moving toward deeper safety

The US researchers propose several solutions, including training models with “safety recovery examples”. These would teach models to stop and refuse even after beginning to produce harmful content.

They also suggest constraining how much the AI can deviate from safe responses during fine-tuning for specific tasks. However, these are just first steps.

As AI systems become more powerful, we will need robust, multi-layered safety measures operating throughout response generation. Regular testing for new techniques to bypass safety measures is essential.

Also essential is transparency from AI companies about safety weaknesses. We also need public awareness that current safety measures are far from foolproof.

AI developers are actively working on solutions such as constitutional AI training. This process aims to instil models with deeper principles about harm, rather than just surface-level refusal patterns.

However, implementing these fixes requires significant computational resources and model retraining. Any comprehensive solutions will take time to deploy across the AI ecosystem.

The bigger picture

The shallow nature of current AI safeguards isn’t just a technical curiosity. It’s a vulnerability that could reshape how misinformation spreads online.

AI tools are spreading through into our information ecosystem, from news generation to social media content creation. We must ensure their safety measures are more than just skin deep.

The growing body of research on this issue also highlights a broader challenge in AI development. There is a big gap between what models appear to be capable of and what they actually understand.

While these systems can produce remarkably human-like text, they lack contextual understanding and moral reasoning. These would allow them to consistently identify and refuse harmful requests regardless of how they’re phrased.

For now, users and organisations deploying AI systems should be aware that simple prompt engineering can potentially bypass many current safety measures. This knowledge should inform policies around AI use and underscore the need for human oversight in sensitive applications.

As the technology continues to evolve, the race between safety measures and methods to circumvent them will accelerate. Robust, deep safety measures are important not just for technicians – but for all of society.

The Conversation

Lin Tian receives funding from the Advanced Strategic Capabilities Accelerator (ASCA) and the Defence Innovation Network.

Marian-Andrei Rizoiu receives funding from the Advanced Strategic Capabilities Accelerator (ASCA), the Australian Department of Home Affairs, Commonwealth of Australia as represented by the Defence Science and Technology Group of the Department of Defence, and the Defence Innovation Network.

ref. How we tricked AI chatbots into creating misinformation, despite ‘safety’ measures – https://theconversation.com/how-we-tricked-ai-chatbots-into-creating-misinformation-despite-safety-measures-264184