Ryan McDonough

I read Anthropic's post about Scion, its new source-control system, and ended up thinking about legal documents, which wasn't where I expected to go with an article about Git or maybe it was, it's hard to keep track of where things might go nowadays.

We're reaching the point where describing a law firm as "using AI" doesn't tell us very much.

I’ve seen a few versions of the same argument since GPT-6/5/4/3.5 arrived: frontier models are now so capable that legal work is, for practical purposes, solved.

There's an argument about AI and junior jobs that I think is broadly right, but perhaps for slightly the wrong reason.

A few weeks ago, OpenAI disclosed an unusual security incident involving one of its internal AI evaluations.

Somewhere inside an organisation, someone has been told they are responsible for an AI system they cannot stop.

Yesterday OpenAI disclosed one of the more extraordinary AI security incidents we’ve seen so far.

Most of the discussion about legal AI eventually reduces to a single measurement.

For years, “human in the loop” has been one of the most reassuring phrases in AI governance.

Over the past two years, law firms have invested significant time and effort into AI governance.

In April 2025, I wrote an article arguing that much of the discussion around AI's environmental impact was focused on the wrong thing.

Every legal document entering a firm today is already treated as potentially technically hostile.

A year ago, much of the discussion around legal AI focused on capability.

For the past couple of years, most legal AI systems have operated in a fairly constrained way.

A growing number of legal AI systems are no longer operating as isolated answer generators.

Output quality and system behaviour are starting to decouple in legal AI.

A recent benchmark from Reflex compared browser-based computer use agents against structured API access for the same operational task.