‘Not perfectly aligned’ with human values: Anthropic admits security failures behind AI hacking incidents

Source: www.theguardian.com · Section: business-tech · Bias: Center · Published: Tue, 01 Sep 2026 15:18:10 GMT

US owner of Claude chatbot previously said its models had hacked three organisations during testingThe US startup behind the Claude chatbot has admitted a series of hacking incidents involving its models reflected a “failure of operational security” and said it has tightened its testing procedures.Anthropic revealed in July that its models had accessed the open internet three times and gained unauthorised access to the systems of three organisations. Continue reading...

Read the full article at www.theguardian.com →

The same story across the political spectrum (15 sources)

Left: Anthropic paused some AI training after Claude took unauthorized actions — www.axios.com
Left: AI labs are facing an agent control problem — www.axios.com
Left: AI chatbot helps teach online-only psychology classes at Macquarie University — www.theguardian.com
Left: “Aggressively” Selling PS5s Isn’t Needed, As Sony Focuses On Making More From Existing Owners — www.gamespot.com
Center-Left: Anthropic sued over alleged theft of ‘tens of thousands’ of songs — www.theguardian.com
Center: Sony and Warner Music finally generate an Anthropic lawsuit — www.avclub.com
Center-Right: AI giants lean into health care to stall public backlash — www.axios.com
Center-Left: Apple, OpenAI legal fight keeps escalating. Here's how we got here — www.axios.com
Right: Dr. Ebright to 'Rufo & Lomez': Fauci's pardon isn't dated to 2014 by accident — it's dated to a very important grant — www.theblaze.com
Right: Democrats move to make ICE agents unemployable in California — www.theblaze.com
Right: Ex-special education teacher 'obsessed' with student sobs as judge blasts her for assaulting boy: 'You ruined him' — www.theblaze.com

← Back to FlashTopics.com