As a startup founder, investor, or marketing leader, you’re likely focused on the growing impact of artificial intelligence. The state of AI in 2025 indicates a significant transformation in business operations, are you ready for it?
Over 75% of respondents in a recent McKinsey study reported their organizations utilize AI in at least one business function. The use of generative AI, in particular, is rapidly increasing. What does this mean for you and the future of your work?
Table of Contents:
- Companies Embracing AI Across Departments
- How Leaders Are Organizing Gen AI Deployment
- Gen-AI-Related Risks
- Companies Following Best Practices in AI Adoption
- The Shifting Skill Needs of AI
- How Businesses Function
- Gen AI Outputs and Review
- Agentic AI Development by 2025
- AI Applications in the Workplace
- The Future of Multi-Modal AI Usage by 2025
- Conclusion
Companies Embracing AI Across Departments
AI use is expanding across various departments within companies. It’s no longer confined to isolated areas of the business. Most companies now utilize AI in multiple functions, such as IT and marketing.
Organizations report using AI in an average of three business functions. For many, the potential applications and possibilities have been extensively considered.
AI in IT Departments is Skyrocketing
IT departments have experienced a substantial surge in AI usage. Growth in this area increased from 27% to 36% within a six-month analysis period.
The emphasis is often on practical application, identifying where AI can maximize efficiency, productivity, and overall impact. With AI tools, IT departments are able to make a measurable impact.
How Leaders Are Organizing Gen AI Deployment
The organization of AI efforts within a company significantly influences its financial outcomes. Many leaders are already aware of this, whether they manage their own organizations or report to executives.
CEO oversight of AI governance is a key factor associated with improved reported impact. Research indicates that CEO oversight has the most substantial influence on a company’s profit and loss.
Centralized Models for Deploying AI
Data governance in organizations is often managed through a centralized system. Centers of excellence are frequently employed. However, regarding staffing, resources are often both internal and distributed across various units.
AI strategy follows a similar pattern. Approximately 46% of companies report using fully centralized systems, while 39% opt for a hybrid approach.
Differences often arise due to company budget constraints. It is very common to see at least a hybrid approach, where an organization may implement both strategies depending on what best aligns with their objectives and goals.
Gen-AI-Related Risks
A significant majority of organizations are actively working to mitigate Gen AI-related risks. In 2024, many struggled to address the various risks posed by the latest technologies.
Larger organizations report managing a greater number of potential cybersecurity and privacy-related issues. However, the management of output accuracy remains relatively consistent with previous levels.
Addressing Key Gen AI Issues
McKinsey data indicates that many organizations are dedicating time to address inaccuracies in results. They are actively working to resolve issues related to inaccuracy, intellectual property, and general privacy.
| Issue | Mitigation Rate July 2024 | Mitigation Rate Mar-Apr 2023 |
|---|---|---|
| Inaccuracy | 44% | 34% |
| Cybersecurity | 38% | 32% |
| Intellectual Property Infringement | 38% | 28% |
As AI tools expand in the cloud, Wiz highlights the dual nature of this technological advancement. AI promotes speed and innovation in their study on AI cloud usage, revolutionizing traditional processes. However, there’s a need for improved long-term tracking and oversight.
Companies Following Best Practices in AI Adoption
Many organizations are still not implementing the necessary practices to achieve substantial impacts. McKinsey discovered this in one of their research pieces.
Tracking clear KPIs for AI solutions influences the bottom line. In larger organizations, having a well-defined path also contributes positively.
Tracking Key Performance Indicators
Only 20% of organizations reported tracking KPIs, indicating significant untapped potential for long-term growth. It’s crucial to implement automated evaluation tools to track and improve these KPIs efficiently.
Respondents in the research study are more likely to adhere to at least some of the recommended practices in 2024 and heading into 2025. This demonstrates that, like any endeavor, best practices must be followed to drive change, and strategic goals must be integrated into organizational objectives and mission statements.
The Shifting Skill Needs of AI
Companies still face challenges in assembling the right teams to effectively manage the evolving landscape of AI. A major factor is that talent perceives significant changes, according to recent McKinsey research.
One role experiencing substantial growth is that of the AI Data Scientist. Companies are continually seeking to hire individuals for this critical function to enhance their internal impact.
Hiring Trends and Workforce Reskilling in AI
Many study respondents also note that companies have reallocated portions of their workforce. Instead of eliminating roles, they are shifting them internally, and this trend is expected to persist through 2025 and beyond.
This involves learning new methods to integrate into existing workflows, without altering the headcount. Using AI development can involve a process, so getting ahead of this with employees helps.
How Businesses Function
Respondents anticipate minimal changes in workforce sizes in the coming years, even into the state of AI 2025.
For example, those in the financial services sector predict potential reductions. This sector’s outlook differs significantly from that of a marketing firm, which foresees new roles emerging around these evolving AI tools.
Impact on Workforce Size
However, a large number of respondents predict minimal overall changes, for better or worse. The state of AI 2025, in this context, remains status quo.
While some aspects will change, many industries expect minimal major internal disruptions. With the evolution of language models, we will learn more each passing year.
Gen AI Outputs and Review
Companies primarily use GenAI to produce texts, but organizations are experimenting extensively. Respondents working for automotive-related organizations utilized AI for images more frequently than those in other lines of business.
27% indicated that all items created are reviewed before being released externally.
Variety of Gen AI Outputs
Over one-third of organizations, as shown in studies, now generate various forms of imaging, including computer code. This indicates diverse and intriguing applications.
Outputs demonstrate significant growth as they evolve beyond text. New AI systems continue to impress.
Agentic AI Development by 2025
By 2025, a significant focus for AI development will be agentic AI. This involves transitioning from asking AI simple questions to creating programs that automatically complete various tasks.
Consider automation tools. Or software performing tasks currently requiring human intervention. Vellum’s research anticipates that agentic systems will provide new insights and automation, particularly in handling large datasets and complex analytics.
Opportunities of Agentic Workflows
Agentic systems utilize various technologies. These workflows transform how individuals approach problem-solving in the workplace, promoting improved efficiency and reduced manual intervention.
Forbes highlights that AI will impact areas such as mental healthcare through industry innovators. It will revolutionize healthcare in both significant and subtle ways going forward, and much of that involves AI data.
AI Applications in the Workplace
In 2024, AI transitioned decisively from theory to practice. Many applications and workflows now commonly utilize GenAI, as expected by both internal teams and customer bases.
Gen AI enables programs to provide satisfactory responses to human requests, ensuring accuracy, helpfulness, usefulness, and resourcefulness. With appropriate human oversight, outcomes continue to positively impact business applications.
Top AI Applications Being Built
Several prevalent models are being leveraged by AI developers. Some of the most common include: Document analysis, customer service chatbots, natural language outputs, and code automation.
Top AI developers were recently recognized for their contributions to this growing sector. Examples of their work can be seen at recent AI summits.
The Future of Multi-Modal AI Usage by 2025
By 2025, it’s highly realistic, and even anticipated, that businesses and applications will integrate text and files. As tools advance, this integration is likely to accelerate even further.
We now live in a world where systems perform a variety of functions rather than offering one-dimensional output. Even the way international finance news reports facts could be influenced by changes in multi-modal functionality, impacting global business use cases, even in simple content output and strategy.
Multi-Modal Applications are Growing
Many utilize AI for more than just one form of output. Instead, it manages different forms of media within a single process and strategy. The AI models continue to impress from an engineering perspective.
Conclusion
Movies often present a distorted view of reality. In the near term, AI is more likely to transform workflows than to initiate a takeover that eliminates human involvement.
From enhancing fraud security to improving everyday tools, AI proves beneficial across numerous fronts. AI’s history dates back to the 1960s, according to research, dispelling the notion that it’s a recent development.
The state of AI in 2025 envisions enhancements in our lives as AI becomes significantly more helpful and pervasive. We have already seen some AI innovation, and that’s sure to continue long term.
Scale growth with AI! Get my bestselling book, Lean AI, today!
