July 4, 2025
AI Agents: 4 Stocks at the
Heart of the Revolution
Research / Thoughts From Themes
ShareWhile generative AI apps like ChatGPT and Perplexity are incredibly useful, AI agents are where the real potential lies in artificial intelligence. With this technology – which can autonomously make decisions, plan, and execute complex multi-step tasks – businesses can potentially enhance their productivity significantly.
As interest in this transformative technology rises, certain companies are emerging as key players, innovating rapidly, and winning new customers. In this blog, we'll explore four stocks that stand at the heart of the agentic AI revolution today and appear well-positioned for growth as this powerful technology reshapes industries.
Salesforce
Agentic AI is still a very new technology. However, one company that has already made a name for itself in the space is Salesforce. A customer relationship management (CRM) software business, it has developed an agentic AI product called “Agentforce” with the aim of revolutionizing customer service and sales through intelligent automation. Since the launch of this product in October 2024, the company has had considerable success, signing approximately 8,000 customers1 (around 4,000 of these are paying for the service).
One key factor behind the success of Agentforce is the level of trust that Salesforce has within the corporate world. Today, its software is already embedded in over 150,000 companies worldwide2, meaning that it doesn’t have to start from scratch when selling software to businesses.
Another is the way that Agentforce is integrated with Salesforce’s apps (e.g. Slack), its Data Cloud service, and its Tableau Next service (which provides metadata). It’s this framework – which it refers to as “ADAM” (apps, data, agents, metadata) – that enables it to deliver a complete agentic AI experience for the enterprise.
On Salesforce’s Q1 FY2026 earnings call, CEO Marc Benioff – who believes the market for digital labor could be worth up to $12 trillion3 – highlighted some companies that are currently using Agentforce. One such company was Finnair, which is already having thousands of conversations a week with customers via the service. It’s aiming to automate 80% of customer service queries and reduce new rep onboarding time by 25% with the technology4. Other companies using the technology to enhance their efficiency include PepsiCo, OpenTable, ENGIE, and Singapore Airlines.
It’s worth pointing out that Agentforce is still a very small part of Salesforce’s overall business. So, it’s not a major growth driver at this stage. However, growth is rapid. For Q1, the company reported Data Cloud and AI annual recurring revenue of more than $1 billion, up more than 120% year on year5.
ServiceNow
While Salesforce is having success with its agentic AI solutions today, competition from rival ServiceNow is a threat. It’s a platform-as-a-service (PaaS) software company that provides digital workflow solutions for automating and managing business processes. It launched its agentic AI offering in September 2024 and has also won a lot of customers since launch. As of March 2025, the company had around 1,000 customers6 using the technology, including big-name organizations such as EY, Rolls Royce, and BT Group7.
In ServiceNow’s Q1 FY2025 earnings report8, it told investors that throughout the first quarter, it unveiled “breakthrough agentic AI innovations” to autonomously solve the most complex enterprise challenges. As a result, customers can now access thousands of preconfigured AI agents across CRM, HR, IT, Finance & Supply Chain, and other business areas as well as its “AI Agent Studio” to build fully customized AI agents. One advantage of its offering is that thanks to its “AI Agent Orchestrator” service, teams of specialized AI agents can work together across tasks, systems, and departments to achieve a specific goal. Another is that like Salesforce, the offering is tied to a foundational data offering (“Workflow Data Fabric”), meaning that AI agents can access both structured and unstructured data from various sources.
Note that recently, ServiceNow has been developing agents for specific industries. In Q1, for example, it unveiled AI agents for the telecom industry. Built on Nvidia AI, the initial use cases will help communications service providers (CSPs) autonomously handle common, labor‑intensive tasks in customer service and network operations, speeding up problem resolution and improving the customer experience. If the company continues to develop industry-specific agentic AI applications, the growth potential could be significant.
Microsoft
It is well-known that Microsoft has been rolling out generative AI applications (e.g. Copilot) recently. What isn’t so commonly known, however, is that the company is a major player when it comes to agentic AI.
Microsoft envisions a world in which AI agents operate across individual, organizational, team, and end-to-end business contexts, making decisions and performing tasks on behalf of users or organizations. And recently, it has been taking steps to make this vision a reality, unveiling both pre-built agents and new models to help developers and organizations build and deploy their own agents securely.
At the Microsoft Build 2025 conference in May, the company explained how its “Azure AI Foundry Agent Service” empowers professional developers to orchestrate multiple specialized agents to handle complex tasks. Today, companies like Fujitsu and NTT DATA are using Azure AI Foundry to build and manage AI agents that help to speed up sales processes. Meanwhile, Stanford Health Care is using Microsoft’s healthcare agent orchestrator to build and test AI agents that can speed up the workflow for tumor board preparation.
Businesses can also use “Microsoft 365 Copilot Tuning” to create agents using their own company data, workflows, and processes. These agents have the potential to perform highly accurate, domain-specific tasks securely. For example, a legal firm could create an agent that is able to generate documents aligned with the firm’s expertise and style.
Overall, Microsoft offers a comprehensive suite of agentic AI capabilities today, ranging from out-of-the-box integrations within its Copilot products to powerful developer platforms that enable the building of highly customized AI agents for enterprise-specific needs. So, it could potentially be a major player in the agentic AI revolution.
CrowdStrike
While AI agents are likely to be heavily used by businesses in customer-facing areas such as sales and customer service, they are also likely to be used behind the scenes in cybersecurity. And one company that could potentially benefit here is CrowdStrike. It offers a solution called “Charlotte AI,” which founder and CEO George Kurtz refers to as the company’s agentic security analyst. This is designed to strengthen threat detection and response by bringing intelligent automation and autonomous reasoning to cybersecurity operations, allowing firms to defend against sophisticated cyber threats at machine speed and scale.
The appeal of Charlotte AI is that it can autonomously analyze data, draw conclusions, and, within defined guardrails, take authorized actions. This essentially transforms cybersecurity from reactive to proactive. It’s also designed to think like an experienced security analyst, asking and answering investigative questions to enhance processes. Meanwhile, the software is continuously trained on data from stopped breaches and insights from analysts – this feedback loop allows it to continually get smarter, staying ahead of evolving adversary techniques.
It’s worth noting that Kurtz believes that protecting other companies’ AI agents is a major opportunity for CrowdStrike. This is due to the fact that AI agents typically have access to a lot of company information, including identities and business workflows. Analysts at Morgan Stanley agree with Kurtz. They have stated that AI agents are likely to expand the “surface area of IT to be protected,” increasing demand for robust cybersecurity solutions9.
Investing in AI Agents
The emergence of AI agents marks a significant leap in artificial intelligence, holding immense potential to revolutionize how businesses operate. As we've seen with companies like Salesforce, ServiceNow, Microsoft, and CrowdStrike, the early applications of this technology are already demonstrating impressive results, from automating customer service to bolstering cybersecurity.
However, it is important to remember that agentic AI is still in its nascent stages and therefore it is hard to predict which companies will ultimately dominate the industry in the long run. Given the uncertainty in relation to the long-term picture, a diversified investment approach could be the most prudent strategy today.
Footnotes:
1Yahoo! Finance, Agentforce in 4,000+ Deals: Can Salesforce Monetize the Momentum?, as of June 12, 2025
2Codleo, How many companies use Salesforce? Total customer number in 2025, as of April 28, 2025
3Nasdaq, Marc Benioff: Digital Labor Creating $12 Trillion Opportunity, as of March 14, 2025
4S&P Market Intelligence, Salesforce, Inc. NYSE:CRM FQ1 2026 Earnings Call Transcripts, as of May 28, 2025
5Salesforce, Salesforce Reports Record First Quarter Fiscal 2026 Results Exceeds Guidance Across All Metrics; cRPO up 12% Y/Y, as of May 28, 2025
6ServiceNow, ServiceNow to extend leading agentic AI to every employee for every corner of the business with acquisition of Moveworks, as of March 10, 2025
7Constellation Research, ServiceNow Yokohama release ups AI agent game, as of March 12, 2025
8ServiceNow, ServiceNow Reports First Quarter 2025 Financial Results, as of April 23, 2025
9Yahoo! Finance, ‘A Safe Haven in a Volatile Market’: Morgan Stanley Picks 3 Top Cybersecurity Stocks to Buy, as of March 21, 2025