Surely you saw the report in past week’s FT (May 15th if you’re interested in looking it up)?
A respected global consultancy released a ‘research’ report in which “more that half a dozen of the footnotes… directed to web pages that did not exist or did not contain the information cited…”
How could they be so stupid???
You can sympathise – at least a little – with stories from the last couple of years when overburdened criminal lawyers or representatives were upbraided by judges for such slackness.
In 2025 an immigration barrister was heavily upbraided by Upper Tribunal Judge for submitting “entirely fictitious” cases in an asylum appeal. The lawyer attempted to hide the fact he had used ChatGPT, and the judge referred him to the Bar Standards Board for his lack of honesty and integrity.
Also in 2025, in the High Court, a junior barrister was castigated by the judge after defending a homeless man using entirely fictional cases. The instructing solicitors were ordered to pay thousands in costs and referred to their regulators.
And this year, a lay advocate presenting as a lawyer submitted a skeleton argument to the Bournemouth Family Court containing four AI-invented cases. Despite apologies, the judge ruled it was in the public interest to explicitly name the lay advocate in the published judgment to warn the public about AI-related misconduct.
Our mild sympathy might lie in the fact that junior criminal barristers are typically underpaid and massively overworked at this stage of their legal career.
Not that this is an excuse. Not at all.
And certainly the instructing solicitors – who may have provided much of the background case law citations – have no excuse.
Whatever the circumstances of the professional use of AI – trusting the machine is wrong, wrong, wrong.
The trouble is that our ‘oh so clever’ top executives in business are desperate… DESPERATE… for AI to deliver a silver bullet which will suddenly reduce staff costs.
The only trouble with that it’s pie in the sky.
It won’t.
And to think so is extremely dangerous.
Indeed, to return to our opening story, we have to ask what is happening when a massive global consultancy publishes papers where the AI assistance has clearly NEVER BEEN CHECKED.
I suspect it’s back to those seniors who want to save money and have said “Oh you can just use the AI to write it for you. No need for proper, knowledgeable writers. No need for a copy editor.”
The result of this story coming out is that the world will find it hard to give any credibility to publications for this consultancy for a good long while.
How damaging is that?
How long has it taken the consultancy to build up its reputation for insight and quality. How much has been invested? How many years? How many man hours? How much intellectual graft? All p***ed away in a minute.
Even more horrifying is the fact that this instance is not a one-off. Other top consultancies are mentioned in the same FT report as having committed the same egregious error.
There are two truths.
AI is transformative.
AI is not a silver bullet for cost savings.
TOP LEADERS… BEWARE!
