Artificial Intelligence

Exploring the Competitive Landscape of AI Agents in 2025

By Thomas Morgan

In 2024, Salesforce made some big announcements regarding their AI initiatives. CEO Marc Benioff has made no secret of his commitment to agentic AI and Agentforce. For Benioff, agentic AI is “what AI was meant to be” – instead of having reactive copilots, Benioff believes in having proactive agents that are integrated across all corners of the business.

Salesforce certainly isn’t the first company to invest in agents, but they certainly are maximizing their capabilities in ways others before couldn’t. What Agentforce offers to customers is more than just helpful assistance – it’s a problem solver designed to automate tasks and solve specific actionable business problems without the need for human intervention.

While Salesforce have successfully released Agentforce, many other giants in tech – namely Microsoft, Google, Amazon, and IBM – have also taken steps towards integrating agents into their platforms. It seems as if agents are the new flavor in the tech world, with almost every major tech company either experimenting or committing to agentic AI.

In this article, we’ll take a look at how Agentforce has performed in its early stages, as well as analyze how other tech companies are utilizing agents in their business operations.

Agentforce So Far

Having only been released in September 2024, we’re still very much in the early stages of Agentforce. However, from what we have seen so far, we can expect Salesforce’s new shiny tool to make significant waves in 2025.

Salesforce CEO Marc Benioff made the decision last year to “hard pivot” his company’s AI strategy, transitioning away from copilots and going all-in on agentic AI. So far, this decision has been well received in the market, which has been well reflected in their share price.

Salesforce has already closed 1,000 Agentforce deals, with many suitors already giving positive feedback. For instance, Accenture has used Agentforce to automate its sales processes, while The Adecco Group utilizes Agentforce’s capabilities to automate candidate prequalification and enhance CVs.

Agentforce has also already helped Salesforce improve their stock market valuation, with investors at Bank of America being described as “bullish” on Salesforce’s AI strategy, advising investors to buy stocks as they believe it will deliver strong returns.

This has, of course, also been encouraged by the announcement of Agentforce 2.0, which brings significant advancements to Agentforce’s agentic reasoning and retrieval augmented generation (RAG), scheduled for general release in February 2025.

For Marc Benioff, these improvements mark Salesforce as the prominent leader in the CRM AI race, stating that no competitors can meet what Agentforce brings to the table.

“The demand for Agentforce has been amazing – no other company comes close to offering this complete AI solution for enterprises. We’re seamlessly bringing together AI, data, apps, and automation with humans to reshape how work gets done. Agentforce 2.0 cements our position as the leader in digital labor solutions, allowing any company to build a limitless workforce that can truly transform their business.” Marc Benioff, Salesforce Founder

The excitement radiating from Benioff regarding Agentforce has been contagious (as well as non-stop), with the tool being all the talk in the ecosystem for the past few months. But it’s clear to see that for Salesforce, this is more than just a shallow rebrand of copilots. It’s a carefully crafted vision with a well-developed strategy that can be understood by everyone – from the least technical people to the most. It’s these distinctive reasons that put Salesforce right at the forefront of this “agent revolution”.

READ MORE: Salesforce Agentforce Is Generally Available: Here’s How to Get Started Today

Microsoft and Open AI: Trouble in Paradise?

As Salesforce keeps pushing forward with its AI agents, Microsoft is also making waves in this space. Since teaming up with OpenAI in 2019, Microsoft has poured billions into building cutting-edge AI tools.

For the past few years, Microsoft’s big focus has been its Copilot tool. Integrated into Office 365, Dynamics 365, and Microsoft Teams, Copilot serves as a smart assistant, helping users with everyday tasks like drafting emails and analyzing data.

In late 2024, Microsoft announced that agents were coming to their 365 platforms. But instead of shifting focus like Salesforce has with its AI strategy, Microsoft is expanding its efforts – they’re continuing to enhance Copilot while carefully weaving agents into their offerings.

These tools are designed to appeal to the widest possible audience, with millions of professionals already using Microsoft’s ecosystem. While Salesforce is heavily focused on AI for customer-facing workflows, Microsoft leans into the sheer scale and versatility of its user base.

As Satya Nadella, CEO of Microsoft, puts it: “We are committed to building AI that amplifies human ingenuity and enables everyone to achieve more.”

Recently, Microsoft announced plans to roll out ten agents for Dynamics 365. Each agent targets a specific business function, including:

  1. Sales Qualification
  2. Customer Service
  3. Financial Reconciliation
  4. Supply Chain Management
  5. Invoice Processing
  6. Order Fulfillment
  7. Inventory Management
  8. Employee Onboarding
  9. Marketing Campaigns
  10. Project Management

However, tensions have been brewing between Microsoft and OpenAI, as OpenAI considers transitioning to a for-profit model – which will limit the returns for investors, including Microsoft. It also means that Microsoft could lose some influence over OpenAI’s direction and strategy. Analysts believe the shift to for-profit won’t fully play out until 2029, but it’s a dynamic worth keeping an eye on in the months ahead. 

Marc Benioff, Salesforce’s CEO, hasn’t been shy about criticizing Microsoft’s AI offerings. He’s even gone as far as calling them “Clippy 2.0,” referencing Microsoft’s infamous virtual assistant from the late ‘90s. Benioff has publicly dismissed Microsoft’s tools as disappointing and a disservice to the broader AI community.

This raises an interesting question: How much of Benioff’s critique is genuine, and how much is competitive posturing? Highlighting perceived flaws in Microsoft’s tools might be a strategic way to shine a spotlight on Agentforce’s strengths. After all, both companies are vying for the same enterprise audience, and each wants to be seen as the leader in innovation.

Despite Benioff’s criticism, Microsoft is doubling down on AI. According to Reuters, the company plans to invest $80B in AI-enabled data centers worldwide in 2025. These facilities will support the training and deployment of its AI products, reinforcing Microsoft’s commitment to the future of AI.

As both companies refine their AI strategies, the rivalry between Salesforce and Microsoft is heating up. Whether Benioff’s comments reflect an honest critique or a calculated move, one thing is certain: 2025 will be a pivotal year for agentic development for Microsoft.

READ MORE: Salesforce 2024 Wrapped: AI Hype, Missed Earnings, Dreamforce, and Agentforce

How Are Other Tech Giants Using Agents?

Aside from Salesforce, other big tech giants – Google, Amazon, and IBM – have all incorporated agents across all of their offerings. Let’s take a closer look at what they’re doing…

Google

Google is making big moves in the world of agentic AI. Their star player is Gemini 2.0, the latest iteration of their AI model, designed for what they’re calling the “agentic era”. This isn’t just another chatbot – Gemini 2.0 is multimodal, meaning it handles text, images, and even audio. 

On the business side, Google Cloud has rolled out Vertex AI Agent Builder, a no-code platform that lets companies create AI agents without needing to write lines of code. These agents, powered by the Gemini model, are built for automating and simplifying complex tasks. 

There’s also Agentspace, Google’s venture into AI-powered enterprise tools, aimed at automating everything from onboarding to generating personalized marketing content.

Google’s experimental projects, like Astra and Mariner, are equally fascinating – Astra is all about real-time, multimodal assistance, while Mariner is exploring how AI agents can independently navigate the web to gather and process information.

Amazon

Working with retail data, Amazon upholds a lot of responsibility when it comes to using customer data and preferences to help them find products they want/need.

So imagine a future where your virtual assistant doesn’t just remind you to buy something – it places the item in your shopping cart and handles the checkout, all while ensuring you’re getting the best deal. This vision is central to Amazon’s approach, leveraging their deep integration of AI and customer data.

One of the standout innovations is Amazon Bedrock Agents, introduced by AWS. These are designed to interact seamlessly with large AI models and connect directly to your organization’s data and workflows. For example, they could manage tasks like booking travel or processing claims, all while maintaining a high level of personalization. 

The main aim is to reduce friction for both businesses and end users by automating repetitive tasks and encouraging smarter decision-making. It’s a leap forward from reactive AI to proactive agents that take care of the details.

On the physical side, Amazon’s warehouses are transforming with autonomous robots like Proteus and Cardinal, which manage tasks such as transporting goods and sorting packages. These robots are effectively agents in a physical world, acting autonomously to reduce human workload and streamline operations.

Source: Amazon

They’re also paving the way for how agents might operate in more customer-facing scenarios, like delivering packages or even assisting in stores.

What’s exciting is Amazon’s potential to combine these developments. With the e-commerce giant’s expertise, it’s easy to imagine shopping agents that autonomously manage your wish list, reorder essentials, and surprise you with personalized recommendations based on your habits.

IBM

IBM has had a mixed journey in the world of AI, with impressive advancements but also notable challenges along the way. On the positive side, IBM is making strides in agentic AI – their watsonx Orchestrate platform, for instance, integrates AI agents with pre-built skills and templates to streamline business processes. These AI agents are designed to handle complex tasks – from HR operations to troubleshooting – and represent IBM’s push to enhance enterprise productivity.

Similar words spoken by Google and Salesforce were echoed by IBM Vice President Parul Mishra, who also believes we are in “a new age of AI led by agentic capabilities.

However, IBM’s track record with AI hasn’t been without its hurdles. One of the most high-profile challenges came from IBM Watson for Oncology – an ambitious project aimed at using AI to recommend personalized cancer treatments.

Although there was a lot of excitement surrounding this development, the system struggled, giving inaccurate recommendations and having difficulties managing complex medical data. These issues, combined with underwhelming results from its Watson Health division, led to the discontinuation of several initiatives by 2023.

This experience revealed the challenge of translating AI’s potential into reliable, real-world applications, especially in fields as intricate as healthcare.

Beyond healthcare, IBM has faced broader challenges with their AI implementation. Engineers working with IBM’s AI tools have reported difficulties in managing complex processes and leveraging the necessary tools effectively. These struggles have underscored the broader complexities of scaling AI technologies across industries for IBM.

Despite these setbacks, IBM continues to push forward, with the aim of having 50% of specific roles in the company fulfilled by AI agents by 2026. It was clear that they needed to make a hard pivot toward agentic AI systems to address these challenges and improve their public image – it will be interesting how they get on throughout the year.

Final Thoughts

It’s almost become unanimously understood across the tech industry that autonomous agents are the next big thing in the AI world and should be the new focus for large-scale enterprises. But we’re still very much in the early stages of agentic development, and there’s no telling if the cracks will start to appear soon.

As these agents have to handle such vast levels of data and information, it will be interesting to see if they can perform for users consistently and remain precise and reliable.

Let us know your thoughts in the comments below!

The Author

Thomas Morgan

Thomas is a Content Editor at Salesforce Ben.

Leave a Reply