Ankur Desai

To Iran now, where several thousand people are thought to have been killed by the Iranian security forces in recent anti-government protests. Many relatives have found it difficult to identify their loved ones who have died. Photos leaked to BBC Verify show the faces of hundreds of people killed in the violent crackdown. The images were displayed in a South Tehran mortuary and were one of the few ways to identify the dead. I got more from Merlin Thomas.

Mikä näyttää olevan lapsen elämä. Tämä kuulostaa erittäin haastavaa ja vaikeaa käsitellä. Ja sinä ja BBC Verify-tyyppi olemme tutkineet antiturvallisuuden protesseja. Miten te teette tätä? Mutta myös se, että internet-kysymys on haastavaa. Se on myös vaikeaa.

The Iranian supreme leader has admitted and publicly acknowledged that thousands have been killed, but he blames the US, Israel and what he calls seditionists for this. BBC verifies Merlin Thomas there. Now a heads up from us. For more depth and analysis of one of our big stories of the day, you can go on YouTube, search for BBC News and then click on the logo and then choose podcast and then the Global News podcast. There's a new story available every weekday.

Sanja, you can hear from outside a court in the Japanese city of Nara, where protesters staged a small demonstration in support of Tetsuya Yamagami, the man convicted of the murder of the former prime minister Shinzo Abe in 2022. Mr. Abe was shot with a homemade gun in Nara city in western Japan while campaigning for his re-election. Our correspondent Shaima Khalil is in Tokyo and is following the story.

Shaima Khalil, reporting from Tokyo. Still to come on this podcast, OpenAI rolls out age prediction on chat GPT for under 18s. But when it comes to guessing your age, will it get it right? If journalism is the first draft of history, what happens if that draft is flawed?

In 1999, four Russian apartment buildings were bombed. Hundreds killed. But even now, we still don't know for sure who did it. It's a mystery that sparked chilling theories. I'm Helena Merriman, and in a new BBC series, I'm talking to the reporters who first covered this story. What did they miss the first time? The History Bureau, Putin and the apartment bombs. Listen on BBC.com or wherever you get your podcasts.

A new report from the United Nations has declared what it describes as the dawn of an era of global water bankruptcy. The report defines this as a situation when long-term water use and pollution have exceeded renewable inflows of water and key parts of the water system can no longer be brought back to previous supply levels.

The study's lead author is Kaveh Madani, director of the UN University's Institute for Water, Environment and Health, also known as the UN's think tank on water. He's been speaking to ANSOI.

Phrases like district-wide crackdown and non-bailable offence aren't usually ones you'd associate with flying a kite. But for nearly 20 years, authorities in parts of India and Pakistan have been putting in tough measures to try and regulate kite flying after a series of deaths. Now, ahead of the Spring Festival Basant in Pakistan, authorities in the city of Lahore have launched a new safety campaign. A Global Affairs correspondent Ambaras Netirajan was an enthusiastic kite flyer as a child. He told me more.

But stopping it happening is proving to be a big problem for the likes of ChatGPT. The chatbots creators have started rolling out their latest tool to try and solve it. The Newsroom's Will Chalk follows AI stories for us and has been looking over this one.

Jos ihmiset eivät käyttäisi ChatGPT, niin se voi olla helppo käsitellä näitä tarinoita, mutta appi sanoo, että se on lähestymässä yli miljoonia käyttäjiä, ja jos otetaan vain yksi osa tutkimusta Yhdysvalloissa, niin 86 prosenttia opiskelijoista, joita koulutusohjelma puhui, sanoi, että he käyttävät AI. Tämä on tietenkin se asia, jota erityisesti nuoret ihmiset integroivat yhdessä elämänsä.

As you said, keeping it safe for under 18s hasn't always been easy and perhaps kind of the biggest, most explosive, shocking cases is the lawsuits that ChatGPT are facing, alleging that it encouraged people, including teenagers, to take their own lives. Now OpenAI, ChatGPT's creator, denies this and these cases are still going through the courts as well.

The company has also repeatedly spoken about the lengths it goes to to keep young people safe, and that brings us to today. So OpenAI, ChatGPT's creators, have started rolling out this new age verification tool. So historically and at the moment, users provide their age, you just simply tell it. You tell it how old you are, it believes you, and then you can use it.

Nyt ChatGPT sanoo, että se alkaa huomioida käyttäjiä ja nähdä, voiko se käyttää muista taitoista. Joten joko asia, jonka käyttäjät käyttävät, tai miten käyttäjät käyttävät sen aikaa. Ja jos se uskoo, että olet yli 18-vuotias, se laittaa lisää turvallisuutta. Ja jos olet itse asiassa vanhempi ja olet kertonut, että olet yli 18-vuotias, koska siitä, miten käyttäjät käyttävät, voit sitten käyttää selfieä tai jotain IDä, jotta voittaisit aikaa ja saadat asian takaisin.

Okei, mutta sitten jotain, mitä ajatellaan kuten AI, miten helppoa on sitten ristiriitaa? No tämä on jatkuva ristiriita, jota ChatGPT käsittelee ympäristön ja turvallisuuden välillä. Joten sinä listasit muutaman ennen, mutta asioita, joita ChatGPT sanoo, että se ristiriitaa, jos ajattelee, että olet yli 18-vuotias.

Graphic violence, viral challenges that could encourage risky behavior, sexual, romantic or violent roleplay, depictions of self-harm and content that promotes things like body shaming. Now these are things that some people would say it's not healthy for anyone to be engaging with AI about, but that's maybe slightly short-sighted because people are using AI for nuanced things, to write fiction, you know.

And if you put too many restrictions in, it can start misinterpreting perfectly innocent requests, you know, write a chapter about a fight or something, as you doing violent content and therefore blocking it. So this is the battle chat GPT has to strike, right? On one hand...

The newsroom's Will Chalk reporting.

And that's all from us for now, but there will be a new edition of the Global News Podcast later. If you want to comment on this episode or the topics covered in it, you can send us an email. The address is globalpodcast at bbc.co.uk. And you can also find us on X at BBC World Service, and you can use the hashtag globalnewspod. This edition was mixed by Jonathan Greer, and the producer was Charles Sanctuary. The editor is Karen Martin, and I'm Ankur Desai. Until next time, goodbye.

If journalism is the first draft of history, what happens if that draft is flawed?