Despite what can only be described as a herculean effort on behalf of the technology vendors who have already poured trillions of dollars into the technology, the miraculous end goal of an Artificial General Intelligence (AGI) failed to materialise this year. What we did get was a slew of enterprise tools that sort of work, mounting cultural resistance (including strikes and legal action from more quarters of the arts and entertainment industries), and vocal criticism leveled at AI’s environmental impact.
It’s not to say that generative artificial intelligence hasn’t generated revenue, or that many executives are excited about the technology’s ability to automate away jobs— uh I mean increase productivity (by automating away jobs), but, as blockchain writer and research Molly White pointed out in April, there’s “a yawning gap” between the reality that “AI tools can be handy for some things” and the narrative that AI companies are presenting (and, she notes, that the media is uncritically reprinting). She adds: “When it comes to the massively harmful ways in which large language models (LLMs) are being developed and trained, the feeble argument that ‘well, they can sometimes be handy…’ doesn’t offer much of a justification.”
Two years of generative AI and what do we have to show for it?
Blood in the Machine author Brian Merchant pointed out in a recent piece for the AI Now Institute that the “frenzy to locate and craft a viable business model” for AI by OpenAI and other companies driving the hype trainaround the technology has created a mixture of ongoing and “highly unresolved issues”. These include disputes over copyright, which Merchant argues threaten the very foundation of the industry.
“If content currently used in AI training models is found to be subject to copyright claims, top VCs investing in AI like Marc Andreessen say it could destroy the nascent industry,” he says. Also, “governments, citizens, and civil society advocates have had little time to prepare adequate policies for mitigating misinformation, AI biases, and economic disruptions caused by AI. Furthermore, the haphazard nature of the AI industry’s rise means that by all appearances, another tech bubble is being rapidly inflated.” Essentially, there has been so much investment so quickly, all based on the reputations of the companies throwing themselves into generative AI — Microsoft, Google, Nvidia, and OpenAI — that Merchant notes: “a crash could prove highly disruptive, and have a ripple effect far beyond Silicon Valley.”
What does 2025 have in store for AI?
Whether or not that’s what 2025 has in store for us — especially given the fact that an incoming Trump presidency and Elon Musk’s self-insertion into the highest levels of government aren’t likely to result in more guardrails and legislation affecting the tech industry — is unclear.
Speaking less broadly, we’re likely to see not only more adoption of generative AI tools in the enterprise sector. As the CIO of a professional services firm told me yesterday, “the vendors are really pushing it and, well, it’s free isn’t it?”. We’re also going to see AI impact the security sector, drive regulatory change, and start to stir up some of the same sanctimonious virtue signalling that was provoked by changing attitudes to sustainability almost a decade ago.
To get a picture of what AI might have in store for the enterprise sector this year, we spoke to 6 executives across several verticals to find out what they think 2025 will bring.
CISOs get ready for Shadow AI
Nataraj Nagaratnam, CTO IBM Cloud Security
“Over the past few years, enterprises have dealt with Shadow IT – the use of non-approved Cloud infrastructure and SaaS applications without the consent of IT teams, which opens the door to potential data breaches or noncompliance.
“Now enterprises are facing a new challenge on the horizon: Shadow AI. Shadow AI has the potential to be an even bigger risk than Shadow IT because it not only impacts security, but also safety.
“The democratisation of AI technology with ChatGPT and OpenAI has widened the scope of employees that have the potential to put sensitive information into a public AI tool. In 2025, it is essential that enterprises act strategically about gaining visibility and retaining control over their employees’ usage of AI. With policies around AI usage and the right hybrid infrastructure in place, enterprises can put themselves in a better position to better manage sensitive data and application usage.”
AI drives a move away from traditional SaaS
Paul Gaskell, Chief Technology Officer at Avantia Law
“In the next 12 months, we will start to see a fundamental shift away from the traditional SaaS model, as businesses’ expectations of what new technologies should do evolve. This is down to two key factors – user experience and quality of output.
“People now expect to be able to ask technology a question and get a response pulled from different sources. This isn’t new, we’ve been doing it with voice assistants for years – AI has just made it much smarter. With the rise of Gen AI, chat interfaces have become increasingly popular versus traditional web applications. This expectation for user experience will mean SaaS providers need to rapidly evolve, or get left behind.
“The current SaaS models on the market can only tackle the lowest dominator problem felt by a broad customer group, and you need to proactively interact with it to get it to work. Even then, it can only do 10% of a workflow. The future will see businesses using a combination of proprietary, open-source, and bought-in models – all feeding a Gen AI-powered interface that allows their teams to run end-to-end processes across multiple workstreams and toolsets.”
AI governance will surge in 2025
“New standards drive ethical, transparent, and accountable AI practices: In 2025, businesses will face escalating demands for AI governance and compliance, with frameworks like the EU AI Act setting the pace for global standards. Compliance with emerging benchmarks such as ISO 42001 will become crucial as organisations are tasked with managing AI risks, eliminating bias, and upholding public trust.
“This shift will require companies to adopt rigorous frameworks for AI risk management, ensuring transparency and accountability in AI-driven decision-making. Regulatory pressures, particularly in high-stakes sectors, will introduce penalties for non-compliance, compelling firms to showcase robust, ethical, and secure AI practices.”
This is the year of “responsible AI”
Mahesh Desai, Head of EMEA public cloud, Rackspace Technology
“This year has seen the adoption of AI skyrocket, with businesses spending an average of $2.5million on the technology. However, legislation such as the EU AI Act has led to heightened scrutiny into how exactly we are using AI, and as a result, we expect 2025 to become the year of Responsible AI.
While we wait for further insight on regulatory implementation, many business leaders will be looking for a way to stay ahead of the curve when it comes to AI adoption and the answer lies in establishing comprehensive AI Operating Models – a set of guidelines for responsible and ethical AI adoption. These frameworks are not just about mitigating risks, but about creating a symbiotic relationship with AI through policies, guardrails, training and governance.
This not only prepares organisations for future domestic and international AI regulations but also positions AI as a co-worker that can empower teams rather than replace them. As AI technology continues to evolve, success belongs to organisations that adapt to the technology as it advances and view AI as the perfect co-worker, albeit one that requires thoughtful, responsible integration”.
AI breaches will fuel cyber threats in 2025
Lewis Duke, SecOps Risk & Threat Intelligence Lead at Trend Micro
“In 2025 – don’t expect the all too familiar issues of skills gaps, budget constraints or compliance to be sidestepped by security teams. Securing local large language models (LLMs) will emerge as a greater concern, however, as more industries and organisations turn to AI to improve operational efficiency. A major breach or vulnerability that’s traced back to AI in the next six to twelve months could be the straw that breaks the camel’s back.
“I’m also expecting to see a large increase in the use of cyber security platforms and, subsequently, integration of AI within those platforms to improve detection rates and improve analyst experience. There will hopefully be a continued investment in zero-trust methodologies as more organisations adopt a risk-based approach and continue to improve their resilience against cyber-attacks. I also expect we will see an increase in organisations adopting 3rd party security resources such as managed SOC/SIEM/XDR/IR services as they look to augment current capabilities.
“Heading into the new year, security teams should maintain a focus on cyber security culture and awareness. It needs to be driven by the top down and stretch far. For example, in addition to raising base security awareness, Incident Response planning and testing
should also be an essential step taken for organisations to stay prepared for cyber incidents in 2025. The key to success will be for security to keep focusing on the basic concepts and foundations of securing an organisation. Asset management, MFA, network
segmentation and well-documented processes will go further to protecting an organisation than the latest “sexy” AI tooling.”
AI will change the banking game in 2025
Alan Jacobson, Chief Data and Analytics Officer at Alteryx
“2024 saw financial services organisations harness the power of AI-powered processes in their decision-making, from using machine learning algorithms to analyse structured data and employing regression techniques to forecast. Next year, I expect that firms will continue to fine-tune these use cases, but also really ramp up their use of unstructured data and advanced LLM technology.
“This will go well beyond building a chatbot to respond to free-form customer enquiries, and instead they’ll be turning to AI to translate unstructured data into structured data. An example here is using LLMs to scan the web for competitive pricing on loans or interest rates and converting this back into structured data tables that can be easily incorporated into existing processes and strategies.
“This is just one of the use cases that will have a profound impact on financial services organisations. But only if they prepare. To unlock the full potential of AI and analytics in 2025, the sector must make education a priority. Employees need to understand how AI works, when to use it, how to critique it and where its limitations lie for the technology to genuinely support business aspirations.
“I would advise firms to focus on exploring use cases that are low risk and high reward, and which can be supported by external data. Summarising large quantities of information from public sources into automated alerts, for example, plays perfectly to the strengths of genAI and doesn’t rely on flawless internal data. Businesses that focus on use cases where data imperfections won’t impede progress will achieve early wins faster, and gain buy-in from employees, setting them up for success as they scale genAI applications.”
- Cybersecurity
- Data & AI
- Sustainability Technology
