State of AI in Business 2025: Why 95% Get Zero P&L
AI
agentic systems
paper
future of work
links
$30–40B invested, 95% of orgs see no P&L impact — five traits the winning systems share.
Despite $30–40B in GenAI investment, 95% of orgs are seeing zero P&L impact. Why? According to the State of AI in Business 2025 report by Project NANDA and MIT researchers [1], it’s because most tools don’t learn, don’t adapt, and don’t integrate.
So what does a winning system look like?
- Agentic by design: Persistent memory + autonomy = fewer prompts, smoother orchestration across workflows.
- Learning-capable: Retains memory, adapts to context, and improves with feedback (unlike static tools).
- Workflow-integrated: Plugs into existing systems with minimal disruption.
- Narrow but deep: Starts with a focused, high-value use case and expands from there.
- Trusted: Clear data boundaries and builder credibility matter more than flashy demos.
References
[1] Project NANDA + MIT. v0.1 State of AI in Business 2025 Report. https://mlq.ai/media/quarterly_decks/v0.1_State_of_AI_in_Business_2025_Report.pdf
Originally posted on LinkedIn.
