AI models may be developing their own ‘survival drive’, researchers say

When HAL 9000, the artificial intelligence supercomputer in Stanley Kubrick’s 2001: A Space Odyssey, works out that the astronauts onboard a mission to Jupiter are planning to shut it down, it plots to kill them in an attempt to survive.Now, in a somewhat less deadly case (so far) of life imitating art, an AI safety research company has said that AI models may be developing their own “survival drive”.After Palisade Research released a paper last month which found that certain advanced AI models appear resistant to being turned off, at times even sabotaging shutdown mechanisms, it wrote an update attempting to clarify why this is – and answer critics who argued that its initial work was flawed.In an update this week, Palisade, which is part of a niche ecosystem of companies trying to evaluate the possibility of AI developing dangerous capabilities, described scenarios it ran in which leading AI models – including Google’s Gemini 2.5, xAI’s Grok 4, and OpenAI’s GPT-o3 and GPT-5 – were given a task, but afterwards given explicit instructions to shut themselves down

‘He’s one of the few politicians who likes crypto’: my day with the UK tech bros hosting Nigel Farage

It is a grey morning in Shadwell, east London. But inside the old shell of Tobacco Dock, the gloom gives way to pulsating neon lights, flashy cars and cryptocurrency chatter.Evangelists for Web3, a vision for the next era of the internet, have descended on the old trading dock to network for two days. For many, the main event is one man: Nigel Farage.“Whether you like me or don’t like me is irrelevant, I’m actually a champion for this space,” the leader of Reform UK tells the audience of largely male crypto fanatics at the Zebu Live conference

‘Sycophantic’ AI chatbots tell users what they want to hear, study shows

Turning to AI chatbots for personal advice poses “insidious risks”, according to a study showing the technology consistently affirms a user’s actions and opinions even when harmful.Scientists said the findings raised urgent concerns over the power of chatbots to distort people’s self-perceptions and make them less willing to patch things up after a row.With chatbots becoming a major source of advice on relationships and other personal issues, they could “reshape social interactions at scale”, the researchers added, calling on developers to address this risk.Myra Cheng, a computer scientist at Stanford University in California, said “social sycophancy” in AI chatbots was a huge problem: “Our key concern is that if models are always affirming people, then this may distort people’s judgments of themselves, their relationships, and the world around them. It can be hard to even realise that models are subtly, or not-so-subtly, reinforcing their existing beliefs, assumptions, and decisions

Meta found in breach of EU law over ‘ineffective’ complaints system for flagging illegal content

Instagram and Facebook have breached EU law by failing to provide users with simple ways to complain or flag illegal content, including child sexual abuse material and terrorist content, the European Commission has said.In a preliminary finding on Friday, the EU’s executive body said Meta, the $1.8tn (£1.4tn) California company that runs Instagram and Facebook, had introduced unnecessary steps in processes for users to submit reports.It said both platforms appeared to use deceptive design – known as “dark patterns” – in the reporting mechanism in a way that could be “confusing and dissuading” to users

Amazon reveals cause of AWS outage that took everything from banks to smart beds offline

Amazon has revealed the cause of this week’s hours-long AWS outage, which took everything from Signal to smart beds offline, was a bug in automation software that had widespread consequences.In a lengthy outline of the cause of the outage published on Thursday, AWS revealed a cascading set of events brought down thousands of sites and applications that host their services with the company.AWS said customers were unable to connect to DynamoDB, its database system where AWS customers store their data, due to “a latent defect within the service’s automated DNS [domain name system] management system”.DynamoDB maintains hundreds of thousands of DNS records. It uses automation to monitor the system to ensure records are updated frequently to ensure additional capacity is added as required, hardware failures are handled and traffic is distributed efficiently

Trump pardons founder of Binance, world’s largest crypto exchange

Donald Trump issued a pardon for the founder of the world’s largest cryptocurrency exchange on Thursday.“President Trump exercised his constitutional authority by issuing a pardon for Mr Zhao, who was prosecuted by the Biden administration in their war on cryptocurrency,” a White House statement said. “The war on crypto is over.”Changpeng Zhao pleaded guilty in late 2023 to one count of failing to maintain an anti-money laundering program and stepped down as CEO of Binance, which paid $4.3bn to settle related allegations