InMarketing Today

Deutsche Bank said today it plans to launch digital asset custody for European corporate and institutional clients this year, pending regulatory clearance. The service will initially support bitcoin, ether and the stablecoins USDC, EURC and EURAU, with the bank managing clients' wallets and private keys through hardware-based key protection, multi-person approvals, and separate warm and cold storage. The initial client base spans corporates, asset managers, hedge funds, custodians, brokers and sovereign institutions served by the bank's Corporate Bank and Investment Bank divisions.

Gerald Podobnik, Deutsche Bank's Co-Head of Corporate Bank, chose his words carefully: digital assets are “not a replacement for the traditional financial system” but a complement to it. For a 150-year-old institution entering a category built on hype and volatility, that restraint is the marketing. Deutsche Bank isn't selling a narrative about disruption; it's publishing a control architecture — key protection, multi-person approval, cold storage — and letting the operational detail do the persuading.

Any brand moving into a category the market still associates with risk should take the same approach: lead with the controls, not the ambition. Publish what you've built to keep the client's assets safe before you publish what you hope the category becomes.

Quartz, a London fintech founded by former Revolut head of global expansion André Silva and former N26 investment product lead Mateus Mesquita Alves, has gone live after raising £2.7 million in seed funding. The round was led by Daphni, with Outward VC and K Fund, and angels including Kantox founder Philippe Gelis. Registered with the FCA as an Account Information Service Provider, the app brings pensions, ISAs, savings and investments into one view and pairs them with Charlie, an AI agent that follows the markets continuously and sends real-time portfolio updates; more than £10 million of members' assets have already been tracked since testing began in the first quarter of this year.

Quartz didn't launch an AI feature; it hired Charlie. Naming the AI agent turns a technical capability into something that reads as a colleague rather than a specification line — a deliberate choice worth noticing alongside co-founder André Silva's own framing: “small differences in how your money works compound into enormous differences in how your life goes.” That sells an outcome, not a product. The by-invitation rollout, after months of testing on real member assets, is its own quiet credibility signal: proof before promise.

If your product includes an AI capability doing real, continuous work on the customer's behalf, give it a name and a personality before you give it a feature list. And hold back the wide launch until you can point to assets already under management, not just a waitlist — scale claims land harder once there's a number behind them.

A new consumer survey from Net Conversion, the Orlando-based impact media agency, finds that 54% of consumers trust a brand less when they can tell its advertising was made by AI, rising to 38% who “strongly agree” among 18-to-29-year-olds. The same consumers are heavy AI users everywhere else: 54% now use AI tools to research purchases, rising to 60% among adults under 40, and 74% of those say AI recommendations widen the number of brands they consider. Just 2% say they're comfortable letting AI complete a purchase on their own.

The split isn't about the technology. Net Conversion chief executive Ryan Fitzgerald put it precisely: “What feels like efficiency inside an agency reads as indifference to the customer.” Consumers already trust AI to do the research on their behalf; what they penalise is a brand that appears to have let AI do the caring on its behalf too. The tell isn't the tool. It's whether the output still looks like someone paid attention.

Keep AI working in the parts of the funnel the customer never sees — research, targeting, personalisation — and keep a visible, accountable human hand on the copy and creative they do see, particularly for under-40 audiences where the trust penalty is sharpest. If AI touches the final asset, budget real time for a human edit pass, because this survey suggests your audience can tell when you didn't.