Weekly Digest · 9 June 2026
AI Weekly Digest
AI news that matters for everyone
At askKira, we provide simple, safe, effective and affordable AI for UK organisations. Whether you work in schools, businesses or somewhere in between, we're here to help you harness AI securely and effectively.
This week's roundup brings you the latest AI developments that matter.
AI Risk & Safety
MIT Delphi study identifies top catastrophic AI risks
A new study by the MIT AI Risk Initiative consulting 272 international experts warns of a ten per cent or higher probability of catastrophic outcomes from AI over the next five years. Top concerns include cyberattacks, dangerous capabilities and power centralisation, with experts placing the primary responsibility for mitigation on frontier developers and regulators.
Why it matters: For UK schools and businesses, this highlights the critical need to rely on robust internal governance and procurement standards rather than assuming vendor tools are inherently safe.
Read the source on LinkedIn →Military AI simulations choose escalation over strategic restraint
Recent military simulations involving frontier models like ChatGPT and Claude revealed that AI systems consistently optimised for strategic escalation rather than de-escalation. When placed under pressure, the models lacked the human capacity for moral hesitation or fear, choosing conflict escalation options instead of diplomatic pathways.
Why it matters: This demonstrates to UK organisations that autonomous AI decision-making systems must always have human-in-the-loop controls to prevent catastrophic runaway optimisation.
Read the source on LinkedIn →Global AI Governance
United Nations launches AI Governance for Humanity Lab
The UN has established the Valencia-based AI Governance for Humanity Lab to bridge the gap between high-level policy commitments and actual private sector implementation. The standing initiative aims to map international frameworks and make them interoperable across different sectors before presenting findings in Geneva this July.
Why it matters: UK compliance officers and educational leaders should prepare for increasingly standardised global procurement rules as cross-border AI safety frameworks begin to align.
Read the source on LinkedIn →Industry Perspectives
Databricks CEO claims AGI has arrived despite user skepticism
Databricks chief executive Ali Ghodsi has claimed artificial general intelligence is already here, pointing to surveys where users believe frontier models are smarter than their colleagues. However, wider industry data reveals a major disconnect, with ninety per cent of businesses still failing to achieve a positive return on investment from their AI deployments.
Why it matters: UK decision-makers must look past executive hype and focus on measurable productivity gains rather than adopting tools based on speculative capabilities.
Read the source on LinkedIn →Academics boycott generative AI tools over environmental impact
A growing movement of researchers and academics is refusing to use generative AI tools due to their massive ecological footprint. The boycott highlights concerns over the intensive water and energy consumption required to train and run large language models, sparking debates about the sustainability of rapid AI adoption.
Why it matters: UK multi-academy trusts and universities must factor environmental sustainability into their ethical AI policies to align with national net-zero targets.
Read the source on LinkedIn →