InMarketing Today

Four in ten UK financial advisers expect demand for their services to rise over the next five years as AI use grows, according to M&G's latest Investor Compass report; a further 39% expect demand to hold steady. Three-quarters of advisers already use AI, mostly to cut administrative time, but their clients lag well behind: 63% have never used AI for their own investment decisions.

The value clients pay for hasn't moved. Peace of mind, not information, is what 56% of investors said was the most valuable part of advice, well ahead of tax planning (15%) and investment management (13%). Ross Liston, chief executive of M&G Advice, put it plainly: "Faster is not always better when people are making major financial decisions." That's a marketing brief as much as a research finding.

Stop marketing information access, because AI has already commoditised it, and start marketing judgement instead. The pitch that wins is a second opinion under pressure and the confidence to decide, not another dashboard.

BNP Paribas has signed a five-year partnership with Google Cloud to deploy agentic AI across its corporate and institutional banking arm, starting with automating parts of corporate credit memo preparation and extending later into sales, trading, research and structuring. Gemini models will also feed into LLM@CIB, the bank's internal AI assistant already used by more than 65,000 employees.

Every line of the announcement reads as though it were written for the compliance officer, not the technologist. Sensitive data stays out of the public cloud, AI agents get authenticated access only to the systems they need, and every interaction is monitored. Charles Holive, BNP Paribas CIB's chief AI officer, frames the whole deal around restraint: "Agentic AI can help our teams handle more complex workflows while maintaining the appropriate level of control." That's the sentence a bank's clients and regulators need to hear before the Google Cloud name-check.

When you announce a major AI vendor deal in a regulated business, lead with what stays under your control, not what the vendor brings. The governance framework is the story; the technology partner is a supporting detail.

AI-written text is everywhere: the Wall Street Journal has run an AI-drafted op-ed, and Pangram, a detection firm, found 41% of longer LinkedIn posts show signs of AI authorship. On quality alone, it can beat humans outright: researchers at Villanova University found readers rated an AI-written short story as "higher quality and more absorbing" than a human-written one in a blind test, The Economist reports.

Quality was never the issue. Trust is. Colleen Kirk, a professor at the New York Institute of Technology, notes that nobody objects to AI drafting a parking-fine appeal, but if a company uses it to apologise for a problematic policy, "you've got a big problem" — the disclosure risks triggering what she calls "moral disgust". Polished AI prose sits in an uncanny valley: competent enough to notice, human enough that the gap still repels.

Match the tool to the moment. Let AI handle the routine and the administrative — nobody cares who drafted the parking appeal — and keep anything carrying an apology, a reputation risk or genuine judgement in human hands. The moment a client suspects the sincerity was outsourced, the content stops working, however well it reads.