Charlotte Gallagher (host / presenter)

My mind went blank and I lost consciousness.

Now recovering at home, Afaf told me she was still in pain, with 20 stitches in her head and bruises on her arms and legs that left her unable to sleep.

The settlers weren't like this at the start of the Gaza war, she told me.

Since then, they've escalated more than in all the years before.

Israeli forces have long been criticised by human rights groups for standing idly by during settler attacks or even taking part in them.

But the level of violence has now triggered warnings from Israel's senior military commanders, who've said it crosses a red line and must be dealt with firmly.

Many hardline settlers see these comments as a betrayal.

At the Hamida Mosque, a group of Israeli activists have arrived to support Palestinian residents here.

I asked Martin Goldberg about Israeli claims that settler violence is overblown.

The head of the Settler Council here issued a statement this week supporting Israeli forces in arresting what he called the anarchists who harmed soldiers and civilians.

But the West Bank's growing security crisis risks exposing dangerous divisions between Israel's military and political leaders.

Extremist settlers say their claim to the land comes from the Bible, but their confidence comes from government support.

That was Lucy Williamson reporting.

An American artificial intelligence company claims one of its programmes has been used by Chinese spies to hack organisations around the world.

Anthropic, said Claude, its rival to ChatGPT and other chatbots, was part of the first reported AI-orchestrated cyber espionage campaign.

But sceptics are questioning the accuracy of that claim, as Joe Tidy explains.

It's the kind of cyber attack that's been feared ever since the rise of generative AI chatbots, the technology that allows us to type in a prompt and get an instant complex response generated as text or even computer code.

Anthropic says it discovered the hack in mid-September and stopped it by cutting off the hackers from their tools.

It banned the unknown individuals and alerted the targeted companies and the authorities.

The company also analysed how its chatbot Claude had been used, and says the hackers from China managed to trick Claude into carrying out many elements of the hacks autonomously.