Author | Lightnews

Resultsense @resultsense.bsky.social · 6h

What do you NOT want AI to do?

AI is fantastic when augmenting your skills and creativity. But for tasks with minimal tolerance for errors, you need to keep it on a firm leash.

Next time you're telling ChatGPT what to do...also tell it what NOT to do.

Resultsense @resultsense.bsky.social · 7h

Shopify Q3: AI traffic 7x, orders 11x since January.

Partners: OpenAI, Perplexity, Microsoft.

64% of shoppers will use AI.

Agentic commerce delivers revenue now.

https://resultsense.com/news/2025-11-05-shopify-ai-traffic-orders-growth/?utm_source=bluesky&utm_campaign=ecommerce&utm_content=1

Resultsense @resultsense.bsky.social · 7h

98% of researchers use AI. 56% save 5+ hours weekly.

But 39% report errors, 37% cite data risks.

Treat AI as junior analyst: fast, needs oversight.

https://resultsense.com/news/2025-11-05-market-researchers-ai-adoption-trust-problem/?utm_source=bluesky&utm_campaign=ai&utm_content=1

Resultsense @resultsense.bsky.social · 9h

42% of UK AI projects fail—not the AI, but the data prep that gets skipped.

Our 5-step framework helps SMEs get data-ready first.

https://resultsense.com/news/2025-11-04-data-readiness-ai-implementation-five-step-framework/?utm_source=bluesky

Resultsense @resultsense.bsky.social · 11h

26% of financial services lack AI compliance controls. Napier AI FCA Supercharged Sandbox: measurable outcomes (£2.54bn target), regulatory partnership. UK AI leadership.

https://www.resultsense.com/news/2025-11-04-napier-ai-fca-supercharged-sandbox/

Resultsense @resultsense.bsky.social · 13h

3-month FCA programme: NVIDIA tech + NayaOne infrastructure. Napier AI testing algorithms on synthetic datasets with 5 Data Science members. Regulatory partnership, not just approval.

https://www.resultsense.com/news/2025-11-04-napier-ai-fca-supercharged-sandbox/

Resultsense @resultsense.bsky.social · 15h

43% of UK SMEs have no AI plans despite £94B opportunity.

Pain → Pilot → ROI → Scale

500% productivity, £20M profit

https://resultsense.co.uk/news/2025-10-27-94-billion-ai-implementation-framework?utm_source=bluesky&utm_medium=organic&utm_campaign=practical-ai&utm_content=framework-post

Resultsense @resultsense.bsky.social · 1d

Big Four precedent: KPMG tracking staff AI engagement now, reviews from 2026 📊

The shift from voluntary exploration to mandatory integration with formal accountability.

What does this mean for UK SMEs?

https://resultsense.com/r/2025-11-03-kpmg-ai-performance-reviews-staff-assessments

Resultsense @resultsense.bsky.social · 1d

FCA Supercharged Sandbox enables AI innovation: £150bn problem → £2.54bn savings. Napier AI frequency-based algorithms with NVIDIA backing. Regulatory partnership throughout development.

https://www.resultsense.com/news/2025-11-04-napier-ai-fca-supercharged-sandbox/

Resultsense @resultsense.bsky.social · 2d

KPMG makes AI adoption mandatory in 2026 staff reviews 🎯

"Everyone has a responsibility to bring AI to their work"—from juniors to leadership.

How should organisations balance AI adoption with workforce support?

https://resultsense.com/r/2025-11-03-kpmg-ai-performance-reviews-staff-assessments

Resultsense @resultsense.bsky.social · 2d

Before deploying AI agents: Map capabilities → Identify failure modes → Establish oversight → Create audit trails → Define accountability. Risk governance transitions from best practice to competitive necessity. https://resultsense.com/r/2025-10-24-companies-agentic-ai-risk-assessment-gap

Resultsense @resultsense.bsky.social · 2d

Salesforce approach:
→ Test internally
→ Prove governance
→ Deploy to customers

Result: 1.8M conversations

95% fail without strategy

Strategic alignment > rushed deployment

https://www.resultsense.com/news/2025-11-01-salesforce-smarter-ai-approach-you-cant-just-throw-llm-at-problem/

1

Resultsense @resultsense.bsky.social · 2d

Salesforce tested Agentforce internally first

Result: 1.8M+ conversations

Now rolling to customers

95% of AI projects fail without strategy

The lesson: prove governance before scale

https://www.resultsense.com/news/2025-11-01-salesforce-smarter-ai-approach-you-cant-just-throw-llm-at-problem/

Resultsense @resultsense.bsky.social · 2d

£180M UK AI investment
+114% skills shortage
Half of tech leaders can't find talent

Tech alone doesn't deliver ROI

Most productive companies: +26% on people

Strategy > Technology

https://www.resultsense.com/news/2025-11-01-from-buzzword-to-bottom-line-4-ways-ai-delivers-roi/

Resultsense @resultsense.bsky.social · 2d

UK AI skills shortage up 114%

Gap: strategy, not technology

4 approaches that deliver ROI:
→ Focus on pain points
→ Align with priorities
→ Invest in people (+26%)
→ Industry solutions

https://www.resultsense.com/news/2025-11-01-from-buzzword-to-bottom-line-4-ways-ai-delivers-roi/

Resultsense @resultsense.bsky.social · 3d

Agentic AI creates new attack surfaces: compromised agents affect infrastructure at scale. Without assessment, orgs miss security vulnerabilities, data protection gaps, operational risks. https://resultsense.com/r/2025-10-24-companies-agentic-ai-risk-assessment-gap

Resultsense @resultsense.bsky.social · 4d

Companies deploy AI agents before risk assessment—autonomous systems making decisions before governance exists. SC Media reveals gap between capability and preparedness. https://resultsense.com/r/2025-10-24-companies-agentic-ai-risk-assessment-gap

Resultsense @resultsense.bsky.social · 6d

The AI recruitment paradox: 73% adoption but 71% believe it removes personalisation. Zinc survey shows automation risks creating barriers to finding the right talent. https://resultsense.com/r/2025-10-23-recruiters-ai-adoption-73-percent

Resultsense @resultsense.bsky.social · 6d

Use AI for admin, not relationships. Free recruiters to build human connections whilst automation handles repetitive tasks. Thoughtful deployment enhances rather than replaces empathy. https://resultsense.com/r/2025-10-23-recruiters-ai-adoption-73-percent

1

Resultsense @resultsense.bsky.social · 6d

73% of UK recruiters now use AI in hiring processes, but 71% believe automation removes personalisation. Zinc survey reveals tension between efficiency and empathy. https://resultsense.com/r/2025-10-23-recruiters-ai-adoption-73-percent

1

Resultsense @resultsense.bsky.social · 7d

Build experimentation capability: Select workflows → Test → Measure → Scale. Coursera: £8K→£16 translations, 4,400 courses in 21 languages. Iteration wins. https://resultsense.com/r/2025-10-24-ai-adoption-quiet-experimentation-middle-ground

1

Resultsense @resultsense.bsky.social · 8d

When AI assistants misrepresent news 45% of the time, users blame both the AI and news sources. EBU study shows error rates threaten journalism trust and democratic participation. https://resultsense.com/r/2025-10-23-ai-assistants-misrepresent-news-45-percent

Resultsense @resultsense.bsky.social · 8d

Verify AI-generated content before publication. Implement source attribution checks. Maintain human oversight. Independent monitoring preserves information integrity. https://resultsense.com/r/2025-10-23-ai-assistants-misrepresent-news-45-percent

Resultsense @resultsense.bsky.social · 8d

45% of AI assistant responses contain significant errors. BBC/EBU study across 14 languages finds systemic failures: Gemini 76% error rate, sourcing problems in 31% of all responses. https://resultsense.com/r/2025-10-23-ai-assistants-misrepresent-news-45-percent

Resultsense @resultsense.bsky.social · 9d

Architecture, not intelligence, defines AI advantage. Success = building coordination enabling humans+AI collaboration. Experimentation > moonshots. https://resultsense.com/r/2025-10-24-ai-adoption-quiet-experimentation-middle-ground

1