9 IT Trends Shaping the Industry in 2025

We’re in the middle of the IT renaissance. According to the latest Future of Jobs Report , demand for roles like AI and machine learning specialists, information security analysts and data engineers is surging globally, with tech roles dominating the top 10 fastest-growing job categories.

However, as opportunity expands, so does the pressure on IT teams to deliver faster and safer systems.

We surveyed more than 2,000 IT decision-makers for our State of IT Security report, 4th Edition, to understand how leaders are navigating these changes. In this article, we’ll unpack the most important insights from that research and provide local examples of it in practice.

Before we dive into the details, here’s a quick snapshot of the top IT trends shaping 2025:

1. Customers want explainability
2. Security budgets are increasing to meet evolving threats
3. Artificial intelligence is both a risk and a security asset
4. Agentic AI is reshaping workforce expectations
5. Cloud security threats are top of mind
6. Compliance complexity is accelerating
7. AI-driven threats are outpacing traditional defences
8. Governance of AI agents is still immature
9. Security transparency builds customer trust

As AI becomes more common in customer experiences (such as interacting with chatbots and getting personalised recommendations), people want to understand how it all works. While accuracy is important to them, explainability is, as well.

Even if your business isn’t actively investing in AI, your employees are likely engaging with AI in the form of Google’s AI overview or using LLMs (large language models) like ChatGPT.

According to our research, 70% of security leaders are concerned about the accuracy and explainability of their AI outputs, but only 43% feel confident explaining them. This dissonance makes it harder to build trust and get buy-in from both customers and internal teams.

If your business is implementing AI, especially in customer-facing tools, it’s important to incorporate explainability from the start. Our Einstein Trust Layer and Data Cloud are designed to help teams keep data secure while also making AI decisions easier to understand and explain.

Security is no longer just an IT responsibility. In fact, we found that 68% of leaders say their security team now plays a more strategic role in the business. As more departments rely on secure computer networks and systems to operate, security is becoming a shared priority.

That’s perhaps one of the reasons 75% of businesses expect their security budgets to increase over the next year. Leaders are recognising that the cost of investing in security now could prevent far greater losses in time, money and reputation if a breach were to happen.

Today’s risks call for a layered approach that’s built into your tech stack from the start. Salesforce Shield 2.0 gives IT teams advanced tools for monitoring, encryption and access control, helping you protect your data without slowing down innovation.

AI is transforming how businesses manage security, but it’s also introducing new concerns. According to our State of IT Security report, 80% of security leaders believe AI agents will bring new security opportunities, yet 79% also say they will create new compliance and security risks.

This dual perspective shows just how complex the AI landscape has become. As tools become more powerful and autonomous, IT leaders are under pressure to balance productivity with protection.

That means building in the right controls from the start so teams can take advantage of AI without exposing the business to unnecessary risk. Salesforce supports this approach with tools like Data Cloud and the Einstein Trust Layer, which help teams manage sensitive data responsibly while still using AI to drive efficiency and insights.

AI agents are becoming more capable and independent. A recent example is OpenAI’s launch of the ChatGPT agent , which can take actions online, such as booking flights, filling out forms and browsing the web.

A Look at ChatGPT agent

Our Agentforce platform was designed to provide businesses with the same level of personal assistance. It can help teams resolve service requests, generate content and deliver intelligent recommendations.

Sutton Tools, an Australian‑owned and operated tool retailer, has already seen strong results with AI agents, including a 20% expected drop in call volumes, hundreds of hours saved each month and a fourfold increase in online orders.

As you can see, having access to these tools can help teams respond to customer requests faster, automate repetitive tasks and make better decisions using real-time data. However, they also change the nature of work. Employees now need to learn how to collaborate with AI and guide its outputs.

If you’re looking to upskill in AI, try our free learning platform, Trailhead. It offers hands-on training to help teams close any digital skill gaps and build their confidence in working with AI tools.

Cloud breaches are now one of the most serious threats facing IT leaders, right alongside ransomware, phishing and data poisoning. With more infrastructure moving to the cloud, even a small misstep in configuration can lead to major exposure.

One example of a business in Australia tackling this problem head-on is SA Power Networks . They’ve launched a five-year cybersecurity plan focusing on who gets access to what, making sure their cloud systems are secure and building security into their tech from the start. They’ve also started using tools to spot weak points in their systems, and they’re getting developers more involved in keeping things safe early on.

As the complexity of threats increases, strong governance, continuous upskilling, and a shared approach to risk are non-negotiable.

As reliance on data increases, governments and industries are stepping up regulations to ensure everyone’s data remains secure. Rules and regulations are changing fast, and many businesses are struggling to keep up. Our research shows that 68% of security leaders say it’s harder to keep up with evolving regulations, and 43% say they feel unprepared for new rules around AI. As laws evolve, companies need to treat compliance as an ongoing process.

In Australia, the government’s AI Ethics Principles lay out clear expectations that AI must be fair, transparent, safe and accountable. These voluntary guidelines are already being used across public sector pilots, and even if you work at a private company, it's worth preparing for them, as they’re likely to shape future regulation.

On the flip side, BizCover is a good example of how businesses can actually use AI to make compliance easier. By integrating Salesforce with an AI-driven solution, the Australian insurance company is able to manage failed payments and support requests for its 200,000 customers, as well as monitor calls and streamline compliance. As a result, they’re meeting their service targets while reducing manual work.

Security leaders see a potential AI-related security problem, as 75% of them have said that AI-powered threats will likely advance faster than traditional cybersecurity tools.

The gravity of this threat became apparent in 2024, when cybercriminals in India used a new form of malware called polymorphic malware (software that constantly changes its code to avoid being detected) to steal over 29 billion rupees . This type of attack is now a growing concern for governments and businesses around the world.

As AI adoption and data volumes grow, the attack surface expands. Cybercriminals are now using AI to scale attacks faster, smarter and at a greater volume, putting more pressure on defenders to modernise their tools.

Today’s risks call for a layered approach that’s built into your tech stack from the start. Salesforce Shield 2.0 gives IT teams advanced tools for monitoring, encryption and access control, helping you protect your data without slowing down innovation.

We found that only 47% of organisations feel fully confident they can deploy AI agents in line with regulatory requirements. In Australia, this is a growing concern as government bodies begin shaping new AI frameworks focused on transparency, safety and accountability.

Without proper oversight, AI agents can introduce legal, ethical and reputational risks, especially in highly regulated sectors like healthcare, finance and insurance.

Using software like Agentforce, with its built-in Einstein Trust Layer, can help businesses use AI safely. It allows teams to set clear guardrails, manage access to sensitive data and monitor how AI agents make decisions. This gives organisations the visibility and control they need to stay compliant as the rules continue to evolve.

In 2023, 52% of customers said they trusted companies less than they did in 2022. In 2024, that number jumped to 71%. As businesses rely more on AI and data to personalise experiences, being open about how that data is used is key to keeping trust.

Concerningly, 41% of CMOs are most worried about data exposure or leakage from generative AI. What’s important in 2025 is to be transparent in your use cases and your commitment to protecting customer data.

Leveraging security and compliance suites like Shield 2.0 can help businesses by giving them more control over how data is accessed, encrypted and monitored. It means they can protect sensitive information and prove to customers that their data is being handled with care.

In 2025, IT leaders are navigating a complex but opportunity-rich environment. With rising workloads, growing AI adoption, stricter regulations and evolving customer expectations, the teams that can adapt have an advantage.

If you want to get more from your software, Salesforce Data Cloud can help you unify your data and deliver intelligent, compliant customer experiences at scale.

Want to learn more about the modern IT industry? Download the full State of IT Security report to explore all the insights from more than 2,000 IT leaders around the world.

You can also find more expert tips, customer stories and thought leadership anytime on Salesforce+.