Microsoft's Security Copilot, Salesforce in Retail, and AI's Role in Workflow Automation

Imagine a world where cyber threats are detected and neutralized before you even know they exist. Welcome to The AI Agent Daily Brief, your go-to for the latest AI updates. Today is Monday, March 24, 2025. Here’s what you need to know about Microsoft's groundbreaking advancements in AI security. Let’s dive in. In a bold move to stay ahead of the ever-evolving cyber threat landscape, Microsoft has unveiled its latest evolution in AI security: the Microsoft Security Copilot agents.

These AI agents are designed to autonomously assist with critical areas such as phishing, data security, and identity management. The pace and complexity of cyberattacks have skyrocketed beyond human capacity, making these AI agents not just a luxury, but a necessity for modern security. Just last year, Microsoft detected over 30 billion phishing emails targeting customers. The sheer volume of these attacks can overwhelm security teams relying on manual defenses.

Microsoft’s new phishing triage agent, part of the Security Copilot, is a game-changer. It can handle routine phishing alerts, freeing up human defenders to focus on more complex threats. This is just one example of how AI agents are transforming the landscape of cybersecurity. But Microsoft is not stopping there. They are expanding Security Copilot with a suite of six new security agents built by Microsoft, and five developed by partners.

These agents are tailored to autonomously manage high-volume security tasks, seamlessly integrating with Microsoft Security solutions. From triaging phishing alerts to optimizing conditional access and remediating vulnerabilities, these AI agents are set to revolutionize the way organizations approach security.

Microsoft’s commitment to AI-driven security is further underscored by their focus on securing and governing AI itself. As organizations rapidly adopt generative AI, the need to secure these technologies becomes urgent. New capabilities are being developed to manage AI security posture across multimodel and multicloud environments, ensuring organizations can protect their AI investments.

In the words of Alexander Stojanovic, Vice President of Microsoft Security AI Applied Research, "This is just the beginning; our security AI research is pushing the boundaries of innovation, and we are eager to continuously bring even greater value to our customers at the speed of AI."

Seventy-five percent of retailers believe artificial intelligence agents will be necessary to compete by 2026. That's according to Salesforce's annual Connected Shoppers Report. This revelation underscores the growing importance of artificial intelligence in the retail sector. As 2025 progresses, retailers are keenly focused on advancing their artificial intelligence initiatives.

These technologies promise to reshape legacy processes like customer service, loyalty programs, and marketing strategies, all in hopes of enticing and exciting shoppers. But here's the catch: while three-quarters of retailers anticipate artificial intelligence agents becoming essential, only a little over half are currently piloting this technology.

Meanwhile, 43 percent are still evaluating potential use cases for autonomous artificial intelligence, and a small 4 percent have no plans to integrate agents at all. Michelle Grant, Director of Retail Strategy and Insights at Salesforce, points out that although many retailers haven't fully matured their use of agents, the development and deployment of these technologies are expected to accelerate rapidly once pilot tests start proving their worth.

Grant notes, "Right now, we’re seeing a lot of deployments, but on a smaller scale. As people test and learn, and once those tests come back, it’s going to be scaled very quickly, and it’ll differ by function." This means not every part of a company will be agent-driven immediately, but certain functions will likely see swift transformation.

The stakes are high, as six in ten retailers express excitement about the potential of autonomous artificial intelligence, with 67 percent seeing it as a business opportunity.

Retailers are eyeing customer service, merchant skills, and marketing as the top three areas where artificial intelligence can make a significant impact. Imagine an agent handling low-value customer returns or drafting consumer-facing marketing emails. These tasks, once human-led, are now on the brink of automation. Consumers are also getting on board.

Seven in ten consumers express interest in using artificial intelligence agents for optimizing loyalty points and managing customer service inquiries. Some are even looking forward to more advanced uses, like having agents buy items when they hit a target price or secure high-demand items before they sell out. As Grant puts it, consumers are tired of constantly monitoring products and just want the friction removed from their shopping experiences.

However, the road to fully implementing artificial intelligence isn't without its hurdles. The cost of integration remains a significant barrier. Despite the clear financial and efficiency benefits, convincing leadership to invest heavily in these initiatives can be challenging, especially with the retail industry's current obstacles, like rising customer acquisition and return costs. Grant emphasizes that part of the struggle lies in disjointed data and existing tech systems.

Although nearly nine in ten retailers have unified commerce initiatives, only 15 percent have realized its value. She stresses that data management must come first, as artificial intelligence relies heavily on high-quality data to function effectively. "Data management has to come first. If you have poor quality data, or you don’t have enough of it, then the artificial intelligence isn’t going to give you results that are accurate or actionable," she explains.

As retailers adapt to incorporate more artificial intelligence, maintaining strong customer service remains crucial, especially as economic uncertainties loom. Salesforce data shows that about three-quarters of shoppers switched brands last year, mainly due to high prices and poor customer service. Grant suggests that loyalty programs could become more vital than ever in maintaining customer engagement.

Consumers agree, with 84 percent stating that loyalty programs increase their likelihood of repeat purchases. However, retailers need to address issues like slow point accumulation and rewards that don't justify the spending. Grant notes that retailers are now reevaluating their loyalty programs more frequently to keep pace with their strategic importance. "Retailers used to update their loyalty programs every three years.

Now, because they’re so strategically important, they’re evaluating them annually," she says.

Oracle is setting the stage for a new era of AI integration within business operations, and it's doing so with a flair for customization. The company has introduced the AI Agent Studio, a groundbreaking platform that lets businesses orchestrate and tailor AI agents within its Fusion Applications suite.

This was announced at the OracleCloud World Tour in London, where Oracle unveiled how companies can now coordinate teams of AI agents to handle tasks across enterprise resource planning, human resources, supply chains, and customer experience systems. Imagine having the capability to customize AI agents to fit seamlessly into your unique business processes. That's exactly what Oracle's AI Agent Studio offers.

Users can modify prebuilt Oracle agents by tweaking their logic, integrating external tools, or even adding custom prompts. Plus, there's the flexibility to choose from a variety of large language models optimized for Oracle or industry-specific needs, like Llama and Cohere. This makes the platform incredibly versatile and adaptable to different business environments. But Oracle's innovation doesn't stop there.

This move builds on their prior AI deployments within their cloud applications, where agents have been embedded to manage routine operations such as invoice processing or recruitment tasks. The AI Agent Studio takes this a step further by enabling these agents to work collaboratively and tackle more complex workflows. This is a significant advancement in how AI can be leveraged to enhance business efficiency and productivity.

Industry giants like Accenture, Deloitte, and PricewaterhouseCoopers are singing Oracle's praises, hailing this as a major leap toward smarter enterprise automation. Analysts are echoing these sentiments, highlighting that Oracle's approach allows businesses to maximize AI efficiency across various departments without incurring additional costs. This could prove to be a powerful advantage in today's fast-paced digital workplace.

The Big Four, those titans of professional services, are racing to redefine how work gets done in their industry. Deloitte, Ernst & Young, PricewaterhouseCoopers, and KPMG are all in on the latest AI wave—agentic AI. Imagine AI agents that don’t just chat but actually make decisions and take action without human input. This is the new frontier they’re exploring, and it promises to shake up how these firms operate.

Last week, Deloitte and Ernst & Young made waves by announcing new platforms developed with Nvidia. Deloitte introduced Zora AI, a platform offering 'intelligent digital workers' that can autonomously handle tasks like expense and invoice management. It’s already being rolled out to Deloitte’s finance team and will soon extend to other areas like procurement and customer services. The goal? To create a digital workforce that fundamentally changes how work is done.

Ernst & Young, on the other hand, is deploying its EY.ai Agentic Platform to 80,000 of its tax division employees. This platform will utilize 150 tax agents to streamline tasks such as data collection and tax compliance. Both firms are touting these platforms as transformative. They’re not just about doing the same work faster; they’re about radically changing business operations.

Jason Girzadas, Deloitte’s US Chief Executive Officer, calls it the beginning of the 'autonomous enterprise era.'

The numbers are compelling. Deloitte claims Zora AI will cut finance team costs by 25% and boost productivity by 40%, freeing up thousands of hours annually. For Ernst & Young, the potential is equally dramatic, with plans to redefine over 30 million tax processes in the next year alone. Raj Sharma, EY’s global managing partner for growth and innovation, emphasizes that having actionable insights from AI could be the key to a business's success or failure.

But it’s not just about the technology. These changes are prompting a reevaluation of business models. At the World Economic Forum, EY’s Sharma suggested that these AI agents might lead to a shift from traditional billing methods to a 'service-as-a-software' model where clients pay based on outcomes rather than hours worked. This could be a game-changer in how these firms engage with clients.

KPMG is also on board, integrating AI agents into their client services, with a focus on customer service, quality, and efficiency. David Rowlands, KPMG’s global head of AI, describes a future where AI agents are 'innovative digital teammates' working seamlessly alongside human colleagues. It’s a vision of a workforce that’s always on, crossing borders and breaking down silos.

PricewaterhouseCoopers is similarly exploring how agentic AI can boost operational efficiency and transform customer experiences. Umang Paw, PwC UK's Chief Technology Officer, noted the importance of building trust and responsible AI from the start. Internally, PwC has already deployed agents for data tasks and customer communications, showcasing the potential of these technologies to revolutionize their operations.

Let's talk about the cutting-edge world of business workflows. Enterprises today are on the hunt for more efficiency and productivity, and large language models are stepping up to the plate in a big way. Imagine not just automating tasks, but having AI agents that understand natural language and make data-driven decisions. These agents are transforming workflow automation, moving beyond the old rule-based systems. So, how exactly do these AI agents work their magic?

They leverage the power of large language models like OpenAI’s GPT-4 Turbo, Anthropic’s Claude, and Google Gemini to interpret business inquiries, classify tasks, and fetch data in real time. These agents don't just sit there waiting for instructions—they actively make decisions and provide recommendations, automating responses and executing workflows with minimal manual input. Here’s where it gets even more exciting.

Unlike traditional automation, which struggles with unstructured data, these AI agents excel thanks to their advanced semantic comprehension. They're now even equipped with multi-modal features, meaning they can handle structured data, voice, and graphics in addition to text. This makes them incredibly dynamic and versatile, ready to tackle a wide range of business procedures. To see how this plays out in a real-world setting, let’s dive into a use case: customer service ticket management.