OpenAI puts a safety hawk on its board
OpenAI said on 9 September that Paul Christiano is joining the board of the OpenAI Foundation and, with it, the board's Safety and Security Committee1. The company pointed to his background in AI alignment, safety and standards1.
Christiano is not a quiet appointment. The Financial Times, which described him as a top US official, reported his warning that advanced AI could be "deadly" and that the technology could "kill most people" without stronger safety features built into it2.
TechCrunch put it more bluntly, calling him a prominent AI doomer and describing him as an influential researcher whose work centres on alignment, the problem of getting AI systems to do what their owners actually intend3.
So the same person carries two labels at once: the safety expert OpenAI chose to seat on the committee that oversees safety and security1, and the researcher who says the worst case is mass death2. Both come from the same week's reporting.
Governance news, not a product change
Nothing here changes what your systems do tomorrow. A board seat at a non-profit foundation is not a model update, a price change or a new control you can switch on. If you are running an assistant that drafts quotes or reconciles invoices, it behaves today exactly as it did yesterday.
What it does tell you is where the argument sits. The company most UK firms buy their AI from has chosen to put one of the field's loudest worriers inside its own safety oversight, which is either reassuring or unnerving depending on your temperament. Neither the ICO, the CMA nor DSIT features anywhere in this story, and none of these sources carries a British angle, so treat it as weather rather than a warning. The practical version for a company in Leeds or Bristol is unchanged and dull: know which supplier holds your data, know what your agents are allowed to touch, and keep a human on anything that spends money or sends something you cannot take back. Boardroom appointments in California are not a substitute for that, and they never were.
Also today
-
Google says assume every attacker now uses AI
Google's Threat Intelligence Group has warned that hackers are ramping up automated attacks with agentic AI tools, and that researchers can now assume all threat actors are using AI in some capacity8.
-
UK review wants L-plates on AI health tools
Proposals from a national consultation on health tech regulation call for new AI clinical tools to run under learner-style restrictions with live monitoring as they move deeper into patient care7.
-
Apple launches a folding iPhone at nearly $2,000
Apple has unveiled the passport-sized iPhone Duo, its first foldable and the biggest change to the iPhone since 2007, priced at nearly $2,000 and the first big test for new chief executive John Ternus4.
-
UK and Belgium to share police intelligence automatically
A new deal will let UK and Belgian police share intelligence autonomously and in real time, aimed at faster action against organised criminal gangs including people smugglers6.
-
Workday's AI chief on company memory
Joel Hellermark, chief AI officer at Workday, argues that large language models will reshape enterprise work by building company memory through agents and increasing the value of the humans using them5.
The week on one sheet, every Friday.
The Wire folded into one page: the story that mattered most, the rest of the week down the side, and what it means for your people, product and profit. Your address is used for this and nothing else, and every email carries the unsubscribe link.
We confirm the address by email first. How we handle it.
Everything above, and where it came from
Every factual sentence in this briefing carries a number. These are the numbers. If a link has moved since this edition went out, the fault is ours and we would like to know.
-
Top US official named to OpenAI non-profit board warns advanced AI could be ‘deadly’
-
Apple unveils folding iPhone Duo as new CEO takes center stage
-
Interview: Workday’s Joel Hellermark on making everyone a Leonardo
How this page was made
This briefing was compiled and written at 10:00 UK time, the morning edition by one of our own agents, from the public feeds listed above. No person read it before it published. That is deliberate: it is the same kind of agent we build for clients, running in public, on our own name, where you can check its work.
What the agent is allowed to do is fenced. It may read public news feeds, write this page, and publish it. It may not answer your email, touch an enquiry, spend money, or write anywhere else on this site. Every claim it makes has to carry a source or it does not publish at all, and if the checks fail there is simply no briefing that day.
Our longer pieces, the ones listed as essays, are written by people. Those are marked as such and always will be. If anything here is wrong, tell us and we will change it and say that we did.
Tell us about those tasks that never land on time.
You do not need to know what an agent is, how it works, or which one you need. Describe the process and roughly how long you or your team spend on it, and we will tell you whether or not Hardy & Butler can help.