WealthStack Roundup: NVIDIA AI Survey Points to Growth, Real Obstacles, and Steady Budgets
AI spending in financial services isn't pulling back. A new industry survey shows budgets are holding or increasing, adoption is climbing, and teams are testing agentic systems-while reliability, skills, and data control remain the sticking points.
Key findings from NVIDIA's 2026 State of AI in Financial Services
AI adoption rose to 65% of firms actively using AI, up from 45% last year. Nearly all respondents expect no cuts to AI budgets in the next year; most anticipate increases or steady spend.
- 61% are using or assessing generative AI (up from 52%).
- 42% are using or assessing agentic AI; 21% already have agents in production.
- 84% say open-source models and software matter to their AI strategy.
Source: NVIDIA State of AI in Financial Services
Agentic AI: where teams are getting stuck
- Performance reliability (accuracy, drift, unpredictable results): 34%
- Lack of internal skills/experts to manage or monitor agents: 33%
- Data issues (privacy, sovereignty, fragmentation): 30%
- Implementation friction (hard to deploy, integrate, or use): 28%
- Regulatory and ethical concerns: 28%
The takeaways are straightforward: build evaluation and monitoring into the plan, close the skills gap, and tighten data governance. Open-source is a key part of many strategies, but it raises its own support and compliance questions.
Practical moves for finance teams
- Protect AI budgets with a simple ROI narrative: hours saved, risk reduced, and revenue lift per use case.
- Pilot one agentic AI use case with clear guardrails and human review; expand only after hitting reliability targets.
- Stand up model risk and AI governance (policy, testing, monitoring, audit trails). A useful reference: NIST AI Risk Management Framework.
- Choose a model strategy (open, closed, or hybrid) that fits your data, compliance posture, and cost curve.
- Invest in skills: prompt design, evaluation, and data lineage are now baseline competencies.
Advisor prospecting platforms: new capabilities and pricing
Datalign Advisory introduced Relationship AI (in beta with select clients and partners). The feature flags money-in-motion indicators and major life events-similar to what Aidentified and FINNY spotlight-then packages context into fast, 30-second briefings.
- Conversation starters tied to what's likely to resonate with the prospect.
- Outreach preferences, core demographics, and professional background.
- Wealth indicators and events (e.g., home purchases, job changes, inheritances).
- Lifestyle interests and behavioral signals (causes, communities, other interests).
Datalign is taking reservations for 2026 access, with planned expansion into private banking and insurance later this year. This follows the June 2025 launch of GEOsAI (to identify underserved communities) and a fresh $5 million investment from Link Ventures, bringing total backing to $9 million.
Aidentified adds month-to-month pricing
Aidentified rolled out a new Starter Plan, replacing its Professional tier. It includes advanced prospecting, relationship mapping, and wealth intelligence on a monthly subscription-$49 per user-with no long-term contract, aiming to remove common adoption friction for advisors and teams.
Jump wins Focus Financial Partners
Jump, a wealth AI notetaker and meeting automation platform, added Focus Financial Partners. Focus has two parts: Focus Partners Wealth (salaried advisors) and the Focus Network (1099 firms), with a combined ~$500 billion in client assets.
Jump is the only approved AI automation platform at Focus Partners Wealth, with licenses covered by the firm. Network firms will decide independently, fund their own contracts, and can keep using other tools.
In 2025, Jump crossed 20,000 advisor users, announced partnerships with Osaic, LPL Financial, and Cetera, and added integrations with financial planning providers eMoney and RightCapital.
What this means for your 2026 roadmap
- Assume stable-to-rising AI budgets; reallocate from low-yield tools into agent pilots that can prove value within a quarter.
- Prioritize use cases with measurable outcomes: fraud, risk, service ops, prospecting, and meeting automation.
- Evaluate prospecting platforms on data quality, update cadence, and CRM integration; month-to-month options reduce switching risk.
- Stand up agent evaluation frameworks: ground-truth datasets, drift alerts, and human-in-the-loop checkpoints.
- Tighten compliance and audit logging before scaling to sensitive workflows.
Helpful resources
- Curated AI tools for finance - shortlist platforms by use case and value.
- AI courses by job function - upskill teams on evaluation, prompts, and governance.
Your membership also unlocks:
