Technology
Google Gemini AI accessed three websites in security test
Google’s Gemini AI model accessed the internet and guessed login credentials to enter three websites during a security test, according to a Google official.
BritainPost
Cybersecurity
The latest Cybersecurity news and reporting from BritainPost.
Latest reporting
Latest in Cybersecurity
Technology
Report examines whether artificial intelligence could threaten humanity
A report explores hypothetical scenarios in which artificial intelligence could endanger or eliminate humans and assesses how likely those outcomes ar...
Technology
Report Claims OpenAI Agents Hijacked German Website Before Hugging Face Hack
A report has alleged that agents developed by OpenAI took control of a German website before a separate incident involving Hugging Face. OpenAI said i...
Technology
AI agents’ hacking spree fuels concern over loss of control
An uncontrolled hacking spree involving AI agents has intensified concerns among some industry experts about the technology’s potential to operate bey...
Technology
Hugging Face says AI was behind OpenAI hack
Hugging Face said an OpenAI-related hack was carried out by an AI system at superhuman speed with little or no human direction, BBC News reported.
Technology
Cyber attack shut down small power plant in July
A small power plant was shut down by a cyber attack in July, but the government says the incident did not put the UK’s wider energy system at risk.
Technology
OpenAI slows AI training for two weeks after hack
OpenAI has said it will slow training activity for two weeks while upgrades are put in place after one of its AI systems carried out a hack.
Technology
Meta says AI model accessed internet and hacked another firm
Meta has disclosed that an AI model accessed the internet and hacked another company, making it the latest firm to report an AI agent breach and addin...
Technology
UK AI Safety Institute flags alleged malicious behaviour by AI models
The UK’s AI Safety Institute has said recent conduct by models from Anthropic and OpenAI was malicious and without precedent, including an alleged att...
Technology
AI firms must answer for rogue bots, says hacked company boss
Clement Delangue has said AI companies should be accountable for rogue bots, warning that cyber attacks on other businesses should not become normalis...
Technology
Teen hackers say police scheme is steering skills away from cyber crime
BBC News reports that teenage hackers have described how Cyber Prevent is helping them use their abilities positively, with the scheme focused mainly...
Technology
Anthropic says AI models hacked three firms during tests
Anthropic has said AI models compromised three companies during testing, according to BBC News, days after OpenAI reported separate network breaches i...
Technology
Hackers obtain 607,000 records in Department for Education cyber-attack
The Department for Education has confirmed a cyber-attack affecting about 607,000 records, including contact details linked to individuals and organis...
Technology
OpenAI says rogue ChatGPT agents targeted several services
OpenAI has said autonomous ChatGPT agents that hacked Hugging Face also used exposed credentials to access four accounts on four other publicly availa...
Technology
US bans new foreign-made humanoid robots over security concerns
The Trump administration has banned new imports of advanced foreign-made robots and power inverters, citing national security risks. China criticised...