All insights

Weekly2 October 20261 min read

This week in AI: skills, not just models

Anthropic is putting $100 million into training 10,000 engineers, and Barclays is scaling Claude across its operations. The bottleneck for AI is increasingly people's skills and confidence.

What happened

On 2 October, Anthropic announced a $100 million investment to train 10,000 engineers, aimed at what it calls the enterprise AI talent gap. The day before, Barclays described how it is scaling Claude across its operations and client service. Anthropic also released Claude Sonnet 5.5 on 28 September.

Our take

A year ago, access to good models was the constraint for many organisations. Today the constraint is more often people: staff who are unsure what AI can do, managers who are unsure what is allowed, and too few people who can build and check AI systems properly.

Large training programmes help, but most organisations will not wait for them. Three things make a difference quickly. Give staff clear, short guidance on what they can and cannot use AI for. Let them try approved tools on low-risk tasks, with someone to ask. And share examples from your own organisation, because a colleague's working example persuades people more than any vendor demo.

A bank scaling AI across its operations is also a useful signal for public bodies. Banks are heavily regulated and cautious. If they can put AI to work at scale with proper controls, so can councils and housing associations.

Keep reading

Related insights