Business · 24 February 2026

AI in Business

For business leaders and professionals

At askKira, we provide secure, personalised enterprise AI for UK businesses. Our platform prevents data leaks, protects IP and client data, and gives your teams AI-powered tools at a fraction of the cost of other enterprise solutions.

This week's roundup brings you the latest AI developments that matter for UK organisations.

Workforce Strategy

Klarna to cut workforce by one-third citing AI efficiency

Fintech giant Klarna has announced plans to reduce its workforce from 3,000 to under 2,000 by 2030. CEO Sebastian Siemiatkowski confirmed the reduction is driven largely by AI adoption, with customer service chatbots already performing the work of 700 full-time agents. The company notes that roles relying on human connection and complex problem-solving will remain safe.

Why it matters: This signals a definitive shift from AI as an augmentation tool to a replacement strategy for operational roles, forcing UK leaders to rethink long-term headcount planning.

Read the source on LinkedIn →

HBR study reveals AI intensifies rather than reduces workload

A new Harvard Business Review study challenges the productivity narrative, finding that AI often expands the number of tasks employees must manage. The research suggests that while AI handles drafting, the burden of verification, editing and integration creates a new layer of 'shadow work' that blurs boundaries between work and downtime.

Why it matters: Implementing AI without adjusting output expectations risks burnout rather than efficiency gains.

Read the source on LinkedIn →

Report highlights gender disparity in AI job displacement

A new Brookings analysis indicates that AI impact will not be monolithic across the workforce. The report warns that women in clerical and administrative roles face disproportionate exposure compared to other demographics. It suggests resilience will depend on 'adaptive capacity'—the ability to move into roles requiring high social intelligence.

Why it matters: UK diversity and inclusion strategies must now include specific upskilling pathways for administrative staff to prevent widening gender gaps.

Read the source on LinkedIn →

Governance & Regulation

Ireland publishes National Digital and AI Strategy 2030

The Irish government has released its 2030 strategy, establishing a dedicated AI Office and setting adoption targets for the public sector. The framework emphasises that AI adoption for SMEs is no longer optional but requires a 'digital wellbeing' mindset. It includes specific provisions for AI literacy campaigns and online safety verification.

Why it matters: As a key trading partner, Ireland's regulatory alignment sets a standard for UK businesses operating across the border.

Read the source on LinkedIn →

US Labor Department releases first AI Literacy Framework

The US Department of Labor has published a federal framework for workforce AI literacy. The guidance moves beyond coding skills to focus on 'mental models', requiring employees to understand pattern-based logic and hallucination risks. It mandates critical thinking as a security control and validates 'prompting' as a core competency.

Why it matters: This framework provides a ready-made template for UK HR directors looking to standardise their own internal AI training competencies.

Read the source on LinkedIn →

Six months remaining until EU AI Act enforcement

Organisations have until August 2026 to comply with the EU AI Act's high-risk requirements or face immediate cessation of non-compliant systems. The regulation demands evidence of human oversight that is captured in real-time, not reconstructed retrospectively. Governance must prove that a specific human had the authority to stop the system at the moment of decision.

Why it matters: UK companies with EU clients must audit their governance 'evidence trail' now, as policy documents alone will not satisfy regulators.

Read the source on LinkedIn →

Shadow AI incidents hit 93% of organisations

New data suggests the vast majority of businesses have experienced a 'Shadow AI' incident where staff use unapproved tools. The ICO and NCSC guidance emphasises management over prohibition, noting that bans simply drive usage underground. Effective governance requires a register of approved tools rather than a blanket firewall block.

Why it matters: Leaders must shift from 'policing' to 'provisioning' secure alternatives to ensure data stays within the corporate perimeter.

Read the source on LinkedIn →

On your radar

Protecting Organisational JudgementAs AI drafts your emails and summaries, it learns your company's internal logic. Ensure your enterprise agreements prevent vendors from training their public models on your team's decision-making patterns to protect your competitive 'sovereignty'.
Runtime StoppabilityFor high-risk AI applications, governance is no longer about paperwork. You must technically prove that a human can intervene and 'stop' the AI execution in real-time. If the 'off switch' isn't immediate and binding, it may fail upcoming compliance checks.
The 'Shadow Work' AuditFollowing the HBR study, conduct a quick audit of your teams using AI. Are they spending more time verifying bad outputs than they saved in drafting? If so, the prompt engineering training is likely insufficient.
Insurability of AIInsurers are moving towards 'deterministic change' models. To get coverage for AI systems, you may soon need to prove that your model's decision-making parameters are 'frozen' for reliance and not constantly evolving in production without review.
NewerAI in Business, 3 March 2026OlderAI in Business, 17 February 2026