Shoosmiths ties second £1m bonus pot to AI fluency

TL;DR:

• Law firm Shoosmiths has committed a second £1m staff bonus pot, this time linked to a four-tier “AI fluency framework” rather than a prompt-count target.
• An earlier £1m incentive asked staff to log one million Microsoft Copilot prompts; the firm hit that four months early, generating two million data points.
• New metrics reward how well staff apply AI — AI Aware through to AI Leaders — a notable shift from rewarding usage to rewarding competence.

Shoosmiths is adding a further £1m to its bonus pool to drive AI adoption, but with a telling change in how it measures success. Its first scheme paid out when employees collectively logged a million Copilot prompts — a target met four months ahead of schedule. The new pot rewards a four-level accreditation ladder, moving the firm from counting prompts to assessing competence.

From prompt volume to fluency

The framework runs from AI Aware (baseline understanding of capabilities and risks) through AI Practitioner and AI Advanced to AI Leaders, who handle strategy and client relationships. Chief executive David Jackson said the firm had “created an AI fluency framework to equip our people with the skills that they will need to thrive in an AI age.”

The internal data is the more interesting part. Personal assistants engaged least with the tools, which Jackson linked to job-security anxiety — prompting reassurance sessions stressing that AI is meant to accelerate growth, not cut headcount. Trainees and associates drove most usage, while partners deliberately used the tools less, preserving their supervisory and quality-assurance roles.

That pattern matters for UK professional-services firms watching adoption stall. It echoes the wider picture in recent banking and enterprise pilot coverage: the bottleneck is rarely the technology, but who feels safe — and skilled enough — to use it.

Looking forward

A prompt count is easy to game; fluency is harder to define and harder to fake. Whether Shoosmiths can credibly grade staff against its four tiers — and whether reassurance is enough to bring anxious roles along — will signal whether structured AI accreditation becomes a template for UK firms or stays a recruitment talking point.