Lloyds Bank to Cut £3.8 Billion Over Four Years. Have You Considered What This Means?

Lloyds Bank has announced a cost reduction of £2 billion (approximately ¥380 billion) over the next four years through the use of AI. At the same time, it plans to invest £1.3 billion (approximately ¥250 billion) in new technologies by 2030.

Can we just brush this off as a story about large corporations?

The answer is NO. This scale of cost reduction changes the price itself. The £3.8 billion saved by Lloyds will go towards shareholder dividends, but that’s not all. It will create a level of “normalcy” with lower fees, faster approvals, and better services that competitors cannot keep up with. Can local credit unions and regional banks compete on the same playing field? How can they compete with a bank that has reduced loan approval times from three days to thirty minutes using AI?

This is not just a story about the banking industry. Similar tectonic shifts are beginning across all sectors.

The Real Danger of “Lower Costs”

Many people misunderstand that large corporations’ AI investments are merely about “efficiency.”

It’s different. This is about changing the cost structure.

For example, a back office that previously operated with 100 people may now run with just 20. The savings from the 80 people no longer needed could amount to hundreds of millions of yen annually. That money can be redirected towards “offensive” strategies—acquiring new customers, lowering prices, expanding services. When large corporations seriously implement AI, not only do costs decrease, but the very rules of competition are rewritten.

In the case of Lloyds, a significant portion of the £2 billion reduction will likely come from labor costs and the automation of business processes. AI-driven credit assessments, chatbot customer service, and automated document processing may seem mundane individually, but together they create a company capable of providing the same services at half the cost.

For SMEs, the threat is not that large corporations will use “amazing AI.” The real threat is that large corporations will acquire “cheap, fast, and accurate operations” and come to take away the space of SMEs in both price and quality.

Okta’s Acquisition of Permiso Signals a New Norm in “AI Security”

In the same week, Okta acquired the AI security startup Permiso for approximately $200 million (around ¥30 billion).

What does Permiso do? The company specializes in threat detection for AI agents and non-human identities (API keys, service accounts, bots, etc.). In other words, it is a company responsible for security in an era where non-human entities access systems.

What does this mean for SMEs?

As we enter an era where AI agents automate business tasks, companies will need to manage not only “who accessed the system” but also “what accessed it.” Large corporations can manage this through platforms like Okta. But what about SMEs?

For SMEs that operate by combining various SaaS tools costing a few thousand yen per month, managing the security of AI agents will become a new cost factor. However, at the same time, as major players like Okta develop this area, there is a possibility that affordable security tools usable by SMEs will emerge within a few years. This is a pattern where large corporations’ investments ultimately lower the prices of tools.

This is a point that should be viewed calmly. Not everything is a threat. Thanks to large corporations investing first, the tools that SMEs can use at lower costs will certainly increase.

The Significance of the New Role of “Forward Deployed Engineer”

Another noteworthy trend is the rapid increase in the role known as “Forward Deployed Engineer” in the AI industry.

What is this? Simply put, it refers to those who implement AI in the field rather than those who create it.

If all that is needed is to develop AI models, researchers and ML engineers suffice. However, to integrate that into actual workflows, enabling field personnel to use it and achieve results, a different skill set is required. These engineers understand technology, know the business, and can mobilize people.

Currently, this role is estimated to have only about 2,000 individuals in the U.S., and large corporations are rushing to hire them. This position, pioneered by Palantir, is said to offer salaries ranging from $200,000 to $400,000 (approximately ¥30 million to ¥60 million).

Can SMEs hire such talent? No, they cannot. No SME can afford to hire an engineer with a salary of ¥60 million.

So, what should they do?

Three Concrete Strategies for SMEs

1. Don’t Hire a “Forward Deployed Engineer”. Buy the System.

Large corporations will hoard specialized talent. SMEs should position themselves as users of the systems and tools created by that talent.

Specifically, no-code and low-code AI workflow tools like Dify, n8n, and Make (formerly Integromat) can be used for monthly fees ranging from a few thousand to tens of thousands of yen. Business automation that previously took months for a forward deployed engineer to custom-build can now be accomplished simply by selecting a template and configuring it.

Automation that used to cost ¥3 million to develop through a systems development company can now be achieved with a tool costing ¥5,000 per month. This “cost cliff” is the point that SMEs should pay attention to.

2. Win in Areas Large Corporations Abandon

The fact that Lloyds Bank is streamlining operations with AI also means that it is cutting out areas that are difficult to streamline with AI.

Complex cases requiring individualized responses, transactions where human relationships matter, and finely-tuned services that are closely tied to local communities. As large corporations push for standardization with AI, the value of “work that cannot be standardized” increases.

The greatest weapon for local SMEs is their proximity to customers. Visible relationships, the agility to act with just a phone call, and the trust that “if I ask that president, something can be done”—these cannot be replaced by AI.

However, it is crucial to clearly distinguish between “work that should be done by humans” and “work that should be left to AI”. Routine tasks like invoice processing, inventory management, and first-level inquiry responses should be left to AI, allowing humans to focus on “work that only humans can do.”

3. Start an “AI Utilization Experiment” for ¥10,000 a Month

The most dangerous thing is to do nothing because “it’s still too early for us.”

ChatGPT Plus costs $20 per month. Claude Pro costs $20 per month. Google Gemini Advanced costs ¥2,900 per month. Automatic meeting minutes, email drafts, data organization, and rough drafts for proposals—there are countless experiments that can be conducted to incorporate AI into business without spending even ¥10,000 a month.

What’s important is not to establish a “perfect AI strategy.” First, touch it, use it, and judge for yourself what “works” and what “doesn’t work.”

In six months or a year, the difference between a company that has never used AI and one that uses it daily will clearly show up in sales and profit figures. That gap will only widen over time.

Don’t Misinterpret the Structure

Finally, I want to make one thing very clear.

While large corporations’ AI investments can bring benefits to SMEs as “leftovers,” with tools becoming cheaper, APIs more user-friendly, and know-how being shared, at the same time, the cost advantages gained by large corporations through AI directly threaten the existing businesses of SMEs.

The £3.8 billion reduction by Lloyds. The $300 million acquisition by Okta. The ¥60 million salary for a forward deployed engineer. These figures indicate that large corporations are serious about AI.

The stance that SMEs should take is clear.

Don’t be afraid. But don’t underestimate it.

There’s no need to do the same things as large corporations. There’s no need to stand on the same playing field. However, the moment you close your eyes and say, “AI has nothing to do with us,” you may find that the very playing field of competition has disappeared.

It’s enough to start with ¥10,000. I hope you try using AI in your business today, even just once. That first step will change your survival odds a year from now.