Tuesday, 7 July 2026
FIFA WC: Belgium shatter co-hosts USA's dream in the last 16
Saturday, 4 July 2026
Father and Son Break Three World Records in 18,000 Mile Cycle Around the World



Friday, 3 July 2026
India inaugurates nuclear-powered hydrogen production facility

_8abe4371.jpg)
_d9caa5bc.jpg)
Wednesday, 1 July 2026
People can learn to spot AI faces – but the clues are no longer obvious
Deepfake faces generated via artificial intelligence (AI) have become so realistic that they routinely fool people, with some research suggesting there may be US$40 billion worth of deepfake-related fraud annually by 2027.
Not only do most people struggle to spot AI faces, but as long ago as 2023 we discovered some AI faces are “hyperreal” – they look more real than actual human faces. We also found people are overconfident they can spot AI faces, with the most confident people making the most errors.
Software-based deepfake detectors do exist, but they can’t really explain the reasons for their detections – and they suffer from serious weaknesses. Some can be fooled simply by converting the image type, such as from png to jpg.
But it turns out most people can learn to spot AI faces with an hour or so of practice. In new research published in PNAS, we show there’s a straightforward way to improve detection of deepfakes, by training people to pick up the tell-tale clues through experience rather than direct instruction.
The difference between human and AI faces
In our early research, we discovered a key difference between AI and human faces. AI faces are hyperaverage.
This means AI faces tend to be more symmetrical, proportional and attractive than human faces. But they’re less expressive and memorable – less likely to stand out in a crowd.
Intriguingly, people can accurately and reliably judge these qualities, but frequently misinterpret the clues. For example, people often think that faces that look a bit odd are AI-generated, when in fact human faces are more likely to have distinctive, unusual features.
Although most people struggle to decide whether a face is AI or real, there is one group who are naturally good at picking up on these clues. So-called super-recognisers, who have exceptional human face perception, seem to be attuned to hyperaverageness, making them better at spotting AI faces.
This made us wonder if, for those of us who aren’t super-recognisers, AI detection abilities can be trained like other forms of perceptual expertise.
Learning to spot AI
In our first study, we invited 45 participants into our lab at the Australian National University, and asked them to rate around 100 faces on six qualities that can be used to tell AI faces apart from real ones: distinctiveness, memorability, proportionality, symmetry, attractiveness and expressiveness.
We didn’t tell participants how these clues might help them distinguish an AI face from a real one – they had to figure that part out for themselves.
We told participants which faces were AI and which were human, but we didn’t tell them that the AI faces were more symmetrical or less expressive, for example. They had to learn these clues through experience rather than direct instruction.
Before and after training, we tested participants’ ability to tell AI faces apart from human ones with new faces that were not used in the training.
Training works
In one test, participants were shown three faces – two human and one AI – and asked to select the face that was AI. On this task, average accuracy doubled from 40% before training to 80% afterwards.
Impressively, all participants improved in their AI detection abilities and several achieved close to 100% accuracy. Participants also became faster and more confident in their correct judgements.
To test the robustness of these findings, the Different Minds Lab at the University of Victoria in Canada conducted a replication of the AI detection training with Canadian participants.
The Canadian lab obtained results that were as strong as those reported in the original Australian study. This shows the training is reliable and can work for different groups of people.
The training was also just as effective when it was administered online rather than in person, which suggests it could be a cost-effective remote intervention in deepfake detection.
A promising start
But this doesn’t mean we’ve solved the AI detection problem. Our training used faces produced with one particular generative AI model, called StyleGAN3.
This is one of the most realistic face generators available, but the technology is advancing rapidly and there are many other models.
Our method has potential to adapt to new models by updating the training images and using multimedia, but we don’t yet have evidence that this will work.
The clues we found for spotting AI faces may shift for other models. And other important questions remain: do the training benefits hold up over time? Is the training effective for people of all ages, including older adults or children?
How to improve your chances of spotting AI faces
If you want to get better at recognising AI-generated faces, looking at a lot of examples is a good start. You can see plenty at websites such as Which Face Is Real or This Person Does Not Exist.
While you’re looking, bear in mind the six key factors we identified:
- how distinctive is the face?
- how memorable is it?
- how proportional is it?
- how symmetrical is it?
- how attractive is it?
- how expressive is it?
This exercise may improve your deepfake radar. But the more important takeaway is that AI deepfakes are improving very quickly – they can easily fool us, even if we think we can spot them.
The clues are no longer obvious: they are not based on specific details but on facial impressions which people form rapidly and naturally, but which can be misleading.
At the same time, there is hope. We have shown it is possible to train people to detect AI faces. By combining our human-centred approach with algorithmic detection, we may yet keep up in this cat-and-mouse game of advancing technology.
Interested in undertaking the AI face detection training? You can register here.![]()
Amy Dawel, Clinical Psychologist and Associate Professor, School of Medicine and Psychology, Australian National University; Eric Mah, Postdoctoral Researcher, Department of Psychology, University of Victoria; Jim Tanaka, Professor of Psychology, University of Victoria, and Tanya George, Research Assistant, Australian National University
This article is republished from The Conversation under a Creative Commons license. Read the original article.
Wednesday, 24 June 2026
Australian musicians hate AI using their songs, but have little legal protection
Andrew Cullen, The University of Melbourne
Music from Kylie Minogue, John Farnham, INXS, Midnight Oil, AC/DC, Tones and I, Gotye, Ben Frost, Nick Cave, Tame Impala, Parkway Drive, The Living End and Vance Joy has been found in a database of 12 million songs used to teach artificial intelligence.
This database, listing songs available on YouTube, is used by AI systems to train the ability to recognise and create music. AI relies entirely on these massive databases, trained on almost everything ever placed on the internet. And Australia’s inclusion in these databases is huge. Kylie alone has 182 songs in just one database.
The volume of Australian music used to train AI has caused significant anger in the Australian music industry, driven by the knowledge that “AI for music creators” platforms such as Suno create as much music as Spotify’s entire catalogue every two weeks.
Dobe Newton, co-writer of folk classic I am Australian and member of the Bushwackers, has music included in the databases. He believes there is “no real ethical nor moral underpinning” to current AI music practices.
Jesse Pattinson of The Delta Riggs is concerned about “the opportunity it will take away from real artists”. Screen composer and APRA board member Caitlin Yeo told me she holds “deep concern for the future of music made by humans for humans”.
She described feeling “violated” when discovering her work in these databases, realising decades of her work had been “hoovered up in a second” to “feed companies offshore that pay no taxes”.
While Australian artists are feeling ripped off, the intersection between copyright law and AI makes proving infringements incredibly difficult.
Copyright and AI
Lawsuits across multiple creative industries have covered how copyrighted books, speeches and even pornography have been used to train AI models.
These actions have led to massive settlements. In 2025, AI company Anthropic paid US$1.5 billion to writers who brought a class action lawsuit over the company’s use of a database of more than 7 million books to train its AI.
But when it comes to proving copyright infringement, the devil is very much in the details, and it may be harder for the impacted musicians to prove infringement than authors.
Crucially, these databases often do not contain any copyrighted material. Rather, they contain instructions on where to download the data from, along with associated information to help AI training.
Legal challenges
The distinction about how the databases are packaged up is not a minor thing.
Previous lawsuits surrounding similar databases found simply listing where copyrighted material can be found is not, by itself, copyright infringement. The infringement only takes place when an AI company uses the data to train its model.
Think of these databases like the map to a safe filled with gold. Having the map itself is fine; stealing the gold is when the law is broken. While we know at least some AI companies have used this data map, and that their training relies upon it, legally establishing use and any copyright infringements is a challenge.
This legal battlefield is complicated by the unique nature of music law.
Copyright protects specific expressions, like a distinct melody or recording, but not a general style. AI developers exploit this loophole. Rather than copying note-for-note, they extract underlying patterns, chord progressions and vocal textures to create a pastiche.
To a creator, this feels like theft. But in the eyes of the law it may just be imitation.
Legal challenges aren’t impossible. The German music royalties society successfully sued chatGPT’s OpenAI. Universal and Sony, representing more than 50% of the music industry, have sued Suno for infringing the copyright of more than 60,000 songs, following other cases against AI giant NVIDIA and Suno.
In response to these lawsuits, Suno described its platform as a fair-use training model, saying “learning is not infringing”.
What’s next?
Since artists became aware of these datasets, the international music industry has rapidly coalesced around several class action lawsuits and lobbying efforts. In Australia, 4,000 artists signed a petition calling on the government to increase protections for artists and today, artists held a press conference in Canberra to discuss AI’s impact.
Tech lobbyists are arguing for exemptions from Australian copyright law. Similarly the Australian Strategic Policy Institute argues that copyright law is a “strategic liability” that increases Australia’s reliance on foreign AI models.
There is a fierce tension between fostering innovation and protecting creative industries. But Australia does not have to choose between the two.
The European Union’s AI act forged an alternative path, to prevent AI companies from hiding copyright infringements from artists and rights holders.
From next August, all AI models accessed from within the EU must declare the source of their training data, and comply with local copyright laws, no matter where the AI was built.
These laws may break the veil of secrecy surrounding the AI data usage, and will significantly increase the likelihood of artists being paid when their work is included in these AI models.
The cultural heartbeat of Australia depends upon supporting creatives. Joining or drawing inspiration from the EU’s AI legislation could help protect artists, and ensure they are fairly represented regarding AI works derived from their labour.
At the very least, as Yeo told me, artists “should see a slice of the pie too”.![]()
Andrew Cullen, Senior Research Fellow, School of Computing and Information Systems, The University of Melbourne
This article is republished from The Conversation under a Creative Commons license. Read the original article.
Tuesday, 23 June 2026
2026 FIFA WC: Record-breaking Messi helps Argentina reach knockouts with 2-0 win over Austria
World Bank slashes global growth forecast, ready to deploy $ 100 b
- The World Bank Group has sharply downgraded its global growth outlook, warning that the escalating conflict in the Middle East will push global economic expansion to its weakest level since the COVID-19 pandemic, as higher energy prices, rising inflation and tighter financial conditions weigh on economies worldwide.
- In its latest Global Economic Prospects report released on Thursday, the World Bank forecast global growth to slow to 2.5% in 2026 from 2.9% in 2025, with forecasts for nearly two-thirds of economies revised downward since January.
- While growth is projected to recover modestly to 2.8% in 2027, it will remain 0.4 percentage points below the average recorded during the 2010s, highlighting the lasting economic scars from successive global shocks.
- The report warned that developing economies continue to face significant challenges, with growth expected to slow to a post-pandemic low of 3.6% in 2026 from 4.4% in 2025 before recovering to 4.2% in 2027.
- “Developing countries have faced a series of challenges over the last decade,” World Bank Group President Ajay Banga said. “The impact differs by country, but the basic test is the same: protect people and preserve stability today, without giving up on growth and jobs tomorrow.”
- According to the report, the closure of the Strait of Hormuz has severely disrupted global energy markets, with Brent crude prices now expected to average $ 94 per barrel in 2026, representing a 36% increase over 2025 levels, assuming major supply disruptions ease by July.
- The Bank also warned of a significant rise in fertiliser prices, with knock-on effects on global food costs. Combined with higher energy prices, these developments are expected to push global inflation to 4% in 2026 from 3.3% last year.
- The World Bank cautioned that risks remain heavily tilted to the downside. In a more severe scenario involving prolonged energy supply disruptions and financial market stress, global growth could slow to just 1.3% in 2026 while inflation could accelerate further to 4.4%.
- Among regions, the Middle East, North Africa, Afghanistan, and Pakistan are expected to suffer the sharpest slowdown, with growth forecast to plunge from 3.9% in 2025 to 1.6% in 2026 before rebounding to 5% in 2027 as reconstruction spending gathers pace and trade flows normalise.
- South Asia is projected to remain the world’s fastest-growing region, although growth is expected to moderate to 6.3% in 2026 from 7% in 2025 before recovering to 6.9% in 2027.
- The report also highlighted growing vulnerabilities stemming from rising public debt burdens. Aggregate government debt across developing economies has climbed from less than 40% of GDP in 2010 to more than 70% of GDP today, increasing borrowing costs and limiting governments’ ability to respond to future crises.
- The World Bank noted that countries with elevated debt burdens face disproportionately higher financing costs, underscoring the importance of restoring fiscal buffers and reducing debt levels to create room for investment in infrastructure, healthcare and education.
- The report also pointed to the challenges facing commodity-exporting economies, which account for roughly two-thirds of developing countries and nearly 90% of low-income nations. While commodity price booms can generate substantial revenue windfalls, much of these gains are often spent rather than saved, leaving countries vulnerable when prices reverse.
- To manage volatility, the World Bank recommended stronger fiscal frameworks, sovereign wealth funds with stabilisation mandates, improved domestic revenue mobilisation and greater economic diversification.
- Against the backdrop of the Middle East crisis, the World Bank announced that it is immediately making available $ 50-60 billion through existing financing instruments, including $ 25 billion in pre-arranged financing, to help countries strengthen social safety nets, support fiscal capacity, and provide liquidity to businesses and farms.
- More than 30 countries are already working with the World Bank under the emergency response framework. If the conflict and its economic fallout persist, the institution said it stands ready to scale up support to between $ 80 billion and $ 100 billion over the next 15 months.
- World Bank Deputy Chief Economist and Prospects Group Director Ayhan Kose said the crisis also presents an opportunity for reform.“The conflict has taken a toll on global activity, but every crisis also brings an opportunity,” he said. “This moment should be used to strengthen policy frameworks, invest in infrastructure, accelerate business-enabling reforms, and mobilise private capital to support job creation at scale,” he said. World Bank slashes global growth forecast, ready to deploy $ 100 b | Daily FT
Monday, 22 June 2026
India’s Zee Entertainment signs World Cup 2026 broadcast deal with FIFA
- FIFA has struck a deal with India’s Zee Entertainment to broadcast the World Cup in the country, ending a months-long standoff over the tournament’s availability in one of the last major markets where rights remained unsold.
- While the financial terms of the package – signed on Monday – were not disclosed, FIFA reportedly sought about $ 100 million for the 2026 and 2030 tournaments before slashing its asking price to $ 60 million.
- The deal gives Zee a toehold in India’s sports broadcast market, where the Reliance-Disney joint venture JioStar holds rights ranging from the Indian Premier League (IPL) cricket tournament to the English Premier League football.
- It covers 39 FIFA events over eight years through 2034, including the Women’s World Cup in 2027, according to a joint statement from FIFA and Zee.
- The agreement came just 10 days before the tournament kicks off on 11 June across the United States, Canada and Mexico.
- Only 14 out of the total 104 World Cup games will begin before midnight for fans in India.
- The final will be held in New Jersey on 19 July, beginning at 19:00 GMT, which will be 12:30 a.m. on 20 July in India. By comparison, 98.4% of matches at the 2018 World Cup started before midnight, and 82.5% at the following edition in Qatar. India’s Zee Entertainment signs World Cup 2026 broadcast deal with FIFA | Daily FT
AI upskilling increases salaries by 150 pc on average in India: Report
Sunday, 21 June 2026
Laugh your way to good health

Friday, 19 June 2026
FIFA WC 2026: Co-hosts Mexico beat Korea 1-0, become first team to qualify for knockouts
Thursday, 18 June 2026
Maruti unveils India’s 1st flex-fuel WagonR capable of running on 100 pc ethanol
Wednesday, 17 June 2026
Iconic Kruger National Park Celebrates 100th Year of Protecting African Wildlife, Including the Big 5


Tuesday, 16 June 2026
Global AI spending expected to surge 47 pc to $2.59 trillion in 2026
Sunday, 14 June 2026
India's media and entertainment industry to rise to Rs 3.3 trillion by 2028: Report
‘It just feels like right time for me to step away’: Williamson on retirement decision
Saturday, 13 June 2026
World Communication Awards 2026: Your chance to celebrate excellence

Coffee, hope, and football: The World Cup’s sleepless return


Mexico City: Actress Salama Hayek is seen before the group A match between Mexico and South Africa at the 2026 FIFA World Cup at Mexico City Stadium in Mexico City, Mexico, June 11, 2026. (Photo: Xinhua via IANS)
Mexico City: Singers Andrea Bocelli (L) and Ejae perform before the group A match between Mexico and South Africa at the 2026 FIFA World Cup at Mexico City Stadium in Mexico City, Mexico, June 11, 2026. (Photo: Xinhua via IANS)
Toronto: People attend the FIFA World Cup 2026 Countdown Concert in Toronto, Canada, on Wednesday, June 10, 2026. Ahead of the FIFA World Cup 2026, the live music celebration was held simultaneously in Toronto, Los Angeles and Mexico City across Canada, the United States and Mexico. (Photo: Xinhua via IANS)




