Having a chit-chat with AI about the 2026 Sector Risk Profile

Following the publication of the Regulator of Social Housing’s Sector Risk Profile 2026 last week, Sue Harvey, Director at Campbell Tickell, put Copilot to the test to see what it could tell us about the changes since 2025.

It’s that time of year, mists and mellow fruitfulness etc and I’ve been chatting with Copilot about the Regulator of Social Housing’s sector risk profile 2026 (SRP26). 

Boards across the social housing sector eagerly await the autumn publication of the SRP to spot gaps and refine their top risks. So do we in the Campbell Tickell risk team: we use it in risk workshops and to support boards, audit committees and risk teams. 

At Campbell Tickell, we always compare each SRP with the previous version. This time I thought I’d try AI — catch up with my younger colleagues and see whether it’s all it’s cracked up to be. 

First off, I asked Copilot for a comparison in the length of the two SRPs. It replied that the word count had increased by 27%, which is a lot! 

Then I asked what the main changes were from the 2025 edition and was somewhat surprised to be told “the 2026 profile is noticeably more interventionist and operational than that of 2025”.  

I needed a bit of persuading, so asked for counts of directive versus softer language. Words denoting an obligation or requirement rose by 63%, compared with the reported 27% increase in the document’s length. But softer language increased too, so the counts alone don’t settle the question: 

In producing this analysis Copilot concluded: “Overall, the 2026 profile appears to use significantly more obligatory language alongside more risk and uncertainty language suggesting a firmer regulatory tone.” Perhaps I should concede on that point? 

Back to the comparison: what’s new? Copilot offered seven items — five relatively technical matters, plus two existential threats in AI and climate adaptation:

1. Supported housing licensing; 

2. Digital telephony switchover; 

3. Rent convergence implementation; 

4. Competence and conduct; 

5. STAIRs; 

6. Artificial Intelligence; 

7. Climate adaptation: 

• • Overheating, 
  • Flooding, 
  • Extreme weather,  
  • Physical climate resilience, and
  • Tenant wellbeing impacts

On AI, Copilot says the regulator sees ‘increased use by landlords in service delivery and decision making’. And that it also identifies risks around ‘accountability, malicious actors, complaints volumes, cyber risk, fraud and misinformation’. 

On climate change Copilot notes a shift of emphasis ‘from carbon reduction (net zero and EPC) to climate adaptation’. After this summer’s string of heatwaves, that feels timely: rebalancing the focus more on mitigation and less on prevention.  

My next question was ‘what risks are mentioned in both SRP25 and SRP26?’, and a long and familiar list was forthcoming.

Copilot identified the following risks:

1. Safety and quality of existing homes 

2. Building new homes 

3. Construction and contractor risk 

4. Housing market risk 

5. Diversification 

6. Financial viability / reduced resilience 

7. Debt and refinancing risk 

8. Fixed-asset sales / disposals  

9. Insurance 

10. Fraud  

11. Data quality / integrity 

12. Governance, risk management and internal controls 

13. Reputation and stakeholder confidence 

More in hope than expectation, I asked what had been dropped. Copilot identified just one topic: ‘the growth of for-profit providers’. It also noted ‘less prominence given to waiting lists’ and a shift in the governance narrative from ‘philosophy’ to ‘compliance, controls, assurance, evidence and execution’. That last point rings true: the fundamentals of risk management always merit repeating. 

Apparently doubting my comprehension skills, Copilot then supplied a punchline: 

‘The biggest strategic message in the 2026 profile is that RSH believes landlords must now deliver on three things simultaneously: 

1. Maintain safe and compliant existing homes. 

2. Increase the supply of new homes. 

3. Remain financially viable despite weak sector finances’

So, my human summary is: more words and a firmer tone, 13 carried-over risks, 7 risks added and 1 dropped. Some very difficult trade-offs, and little by way of low-risk options. Is it just recency bias or is risk management in social housing becoming ever more complex? 

My final experiment? Reading the SRP26 itself! Definitely worth the extra effort for the detail and nuance around individual risks and to fully appreciate the regulator’s particular anxieties. I recommend this old-fashioned approach — though perhaps that’s my age showing. 

Enough, I’m off to speak to some real people!

If you’d like to discuss anything raised in this article, please contact Sue Harvey at sue.harvey@campbelltickell.com

Disclaimer: Please note this article includes AI-generated content that has been reviewed for accuracy.