OpenAI Updates: May 16, 2026

1. ChatGPT launches personal finance tools with Plaid integration

OpenAI. OpenAI released personal finance tools in preview for ChatGPT Pro subscribers in the United States, letting users connect bank, brokerage, and credit card accounts to ask questions about spending analysis, portfolio performance, and financial planning. The product runs on a Plaid integration covering over 12,000 institutions including Schwab, Fidelity, Chase, Robinhood, American Express, and Capital One, with future support for Intuit planned and an underlying GPT-5.5 model tuned for finance reasoning. OpenAI noted that more than 200 million users already ask ChatGPT financial questions monthly, and the feature builds on the April 2026 acquisition of personal finance startup Hiro. Source

2. GPT-5.5 and Codex go live on Databricks under Unity AI Gateway

OpenAI. Databricks announced native availability of GPT-5.5 and Codex across AWS, Azure, and GCP, governed end to end by its Unity AI Gateway for permissions, cost controls, guardrails, and observability. OpenAI positioned GPT-5.5 as its strongest frontier model for agentic enterprise work, complex document reasoning, and long-horizon coding agents, citing a 64.66% score on the OfficeQA Pro LLM benchmark with Oracle PDF and Web Search (up from 57.14% on GPT-5.4) and a 52.63% score on the OfficeQA Pro Agent Harness using Codex (up from 36.10%, a 46% reduction in errors). Source

3. OpenAI Academy publishes Codex playbook for sales teams

OpenAI. OpenAI Academy released a guide showing how sales departments can use Codex to generate pipeline briefs, meeting materials, forecasting reviews, account strategies, and stalled-deal analyses from live CRM and revenue data. The playbook is part of a broader “Codex for Work” series targeting non-engineering functions, repositioning the coding agent as a general-purpose analytical tool for revenue operations and account executives. Source

4. Codex for Work guide targets business operations teams

OpenAI. A companion Academy post details how business operations teams can apply Codex to produce initiative briefs, strategy communications, leadership packets, and progress reports drawn from real operating data. The guide walks through prompt structures and connector setups that let ops analysts compile cross-functional updates without manually stitching together spreadsheets and slide decks. Source

5. Data science teams get a Codex playbook of their own

OpenAI. OpenAI Academy also published a Codex guide for data science teams covering root-cause analyses, impact summaries, KPI memos, scoped assessments, and dashboard specifications. The material targets analytics and ML practitioners who want Codex to handle the documentation and reporting layer around their notebooks and pipelines rather than just code generation. Source