The news: OpenAI’s latest frontier model, GPT-5.4, could boost its competitive edge in financial services and coding and help it become a stronger contender for enterprise adoption in the face of rising competition from Google and Anthropic.

OpenAI called GPT-5.4 its “most capable and efficient frontier model for professional work” and said it’s better at tasks involving spreadsheets, presentations, and documents, with less back and forth with the user. It’s also OpenAI’s most token-efficient reason model to date.

Why it’s worth watching: The latest model is a strategic move in an accelerating race for enterprise AI dominance, where efficiency, automation depth, and cross-department functionality are key assets.

• GPT-5.4’s improvements could push it closer to being a full-stack productivity layer, strengthening OpenAI’s pitch to CIOs as a standardized provider across departments.
• Token efficiency and reduced back-and-forth will lower usage costs and friction—especially important as buyers scrutinize ROI.

OpenAI still faces aggressive competition from Google’s Gemini, which is quickly becoming a top consumer-use model, and Anthropic’s Claude. The latter is gaining market share due to backlash over OpenAI’s Department of Defense (DoD) deal while seeing its finances surge.

• OpenAI surpassed $25 billion in annualized revenues as of February, a source told The Information, up 17% from the end of 2025.
• Anthropic’s annualized revenues recently topped $19 billion, double its results from the end of 2025 and up 36% from two weeks ago.

The bigger picture: If GPT-5.4 performs much better in structured reasoning, tool use, and document handling, marketers could have an easier time creating campaign reports, analyzing performance spreadsheets, building decks, and iterating on messaging. That could speed up adoption among business users.

As professional-grade capabilities improve, consumers may begin to expect similar automation in personal finance, productivity, and creative tasks, which could strengthen ChatGPT’s stickiness as a daily tool.

Recommendations for brands: GPT-5.4 could be the workhorse OpenAI needs right now after lukewarm responses to models like GPT-5, but it’s still a novel tool. Brands should:

• Be wary of single-model dependency. Work with multiple providers to mitigate outages, pricing shifts, or model regressions.
• Plan for unexpected updates. OpenAI’s release speed and sometimes uneven rollouts show how performance and positioning can shift quickly. Train staff on models from multiple providers to make integrations more flexible.
• Experiment within lower-priority pilots. Treat GPT-5.4 as a competitive option in a maturing AI vendor landscape, not a default standard.

This content is part of EMARKETER’s subscription Briefings, where we pair daily updates with data and analysis from forecasts and research reports. Our Briefings prepare you to start your day informed, to provide critical insights in an important meeting, and to understand the context of what’s happening in your industry. Non-clients can click here to get a demo of our full platform and coverage.