Anthropic and Perplexity: AI Duel to Conquer Finance

A Targeted Offensive on Institutional Finance

On May 5, 2026, Anthropic and Perplexity simultaneously launched their artificial intelligence tools dedicated to professional finance, marking a strategic offensive in the institutional market. These two AI giants aim to establish themselves in the institutional finance sector, but with distinct approaches.

Anthropic has deployed a suite of ten AI agents specifically designed to appeal to major financial institutions such as JPMorgan, Goldman Sachs, and Citi. These agents are ready to use and aim to automate the most time-consuming tasks in the financial sector, including pitchbook creation, accounting reconciliation, KYC screening, and monthly closing. Each agent is equipped with specific business instructions, data connectors, and specialized sub-agents, allowing for direct integration into Claude Cowork or Claude Code, or autonomous use via Claude Managed Agents for overnight processing or entire portfolios.

Anthropic's data ecosystem is vast, integrating sources such as FactSet, S&P Capital IQ, MSCI, PitchBook, Morningstar, LSEG, as well as Dun & Bradstreet, Verisk, Third Bridge, and Moody’s. The latter integrates its credit ratings on 600 million companies directly into the Claude interface. The system is based on Claude Opus 4.7, which achieved a score of 64.37% on the Vals AI Finance Agent benchmark, measuring an agent's ability to perform real financial tasks.

Perplexity: Integration as Strength

On the other hand, Perplexity does not target the general public but directly addresses financial institutions such as hedge funds, private equity, and asset managers. Its tool, the "Computer for Professional Finance," integrates 35 dedicated workflows, offering seamless integration with existing tools. The AI connects directly to licensed data terminals, such as Morningstar, PitchBook, and Daloopa, via MCP connectors, or utilizes its 14 integrated providers. The system operates directly within Microsoft Teams and soon in Excel, allowing analysts to launch queries without changing environments.

Perplexity's main argument rests on a clear promise: to eliminate hallucination. In the financial realm, approximations are not tolerated. The system guarantees complete traceability, with each generated figure being clickable and linking back to the exact line of the original SEC document, while also displaying intermediate calculations.

Two Strategies, One Sector Under Pressure

The coincidence of the announcements highlights a convergence: professional finance is seen as a sector where AI can provide immediate and measurable economic value. Jamie Dimon, CEO of JP Morgan, speaks of a profound transformation of work thanks to these tools.

Anthropic and Perplexity are in direct competition, targeting the same investment funds and investment banks, but with different technological philosophies. Anthropic offers an army of customizable agents, suited for foundational work and mass processing in back-office operations, while Perplexity provides a "plug and play" tool, integrated into existing systems, with a particular focus on traceability.

Finance is now a large-scale laboratory for agentic AI. The stakes go beyond the technological framework: they are also psychological and regulatory. How far will an analyst or portfolio manager be willing to let an AI model engage their responsibility and sign on their behalf?