Issue 10
IoD issues ten principles for board oversight of AI; more professional bodies urge board readiness
The Institute of Directors (IoD) published "Introducing AI Governance", which sets out ten principles for board oversight of AI (7 Oct 2026). The same week, Russell Reynolds and the Manufacturers Alliance also pointed to a lack of AI knowledge and governance principles on boards.
Who is affected and how
- igazgatósági tag
- The ten principles are a basis for assessing whether the board's AI knowledge and oversight arrangements meet expectations.
AI erodes the entry-level rung: talent pipeline risks loom for engineers, bankers and HR
Several professional sources say that cutting entry-level headcount in favour of AI could break leadership pipelines (theHRD, Oct. 6, 2026). LeadDev examines the junior engineering pipeline and Gala…
HR-vezetőMap which entry-level tasks AI takes over and how to preserve the learning path regardless.középvezetőSet aside time to review and mentor juniors' work.research · globalGoldman Sachs: AI threatens middle managers as new graduates could do senior-level work from day one
Goldman Sachs warns that, because of AI, new graduates could take on management-level tasks from day one, according to The Economic Times (Oct. 8, 2026). Other pieces describe a silent devaluation of…
középvezetőMiddle-management value is shifting from coordination to judgment, coaching and AI oversight, so this role is worth reshaping deliberately.HR-vezetőA review of management roles, spans of control and promotion criteria is due.survey · globalDeloitte survey: about 80% of organisations lack mature governance for agentic AI
A new Deloitte survey finds that about 80% of the organisations surveyed lack mature governance capabilities for agentic AI, such as clearly defining the scope of action of agents (6 Oct 2026)…
igazgatósági tagBefore agents are deployed, ask for a report on governance maturity.szervezeti döntés · globálisBMW reportedly to cut 20% of management as Apple also builds a flatter structure
According to Autoevolution, BMW would simplify corporate bureaucracy alongside the rise of AI and cut 20% of its management headcount (Oct. 5, 2026). HR Executive says Apple's reported overhaul…
HR-vezetőCommunication of the flattening, reassignment of affected managers and protection of succession pipelines should be planned in advance.középvezetőIt is worth preparing for the requirements of an AI-enhanced role with a wider span of control.Hungary
More key stories of the week
California adds new limits on AI in workplace decisions
California is adding further restrictions on the use of AI in workplace decisions, HR Executive reports (Oct. 5, 2026).
HR-vezetőCompanies need to map which hiring and evaluation decisions use AI and how they comply with the new limits.case · globalHuman–agent teams: managing AI colleagues calls for new leadership skills
Business Insider uses a customer service manager's example to show that managing AI agents can sometimes be more tiresome than managing people (Oct. 6, 2026). Fast Company asked 44 leaders whether AI…
középvezetőDefine agents' tasks, escalation points and accountability within the team clearly.HR-vezetőBuild agents into the system of roles, access rights and performance measurement.case · United StatesThree fired OpenAI safety researchers say they were pushed out in suspicious circumstances; company cites breach of trust
Three OpenAI safety researchers say they were fired in suspicious circumstances (CNN, Oct. 8, 2026). OpenAI stands by its decision and said it was over a breach of trust, not because they raised…
igazgatósági tagThe board should independently verify that risk signals reach it through an open channel.HR-vezetőA separate procedure is needed to distinguish a breach of trust from a safety report.állásfoglalás · globálisChief data and AI officer role struggles to find its place as firms debate who owns AI in the C-suite
CDO Magazine reports that the chief data and AI officer role keeps struggling to establish itself despite growing investment (Oct. 6, 2026). Other pieces debate who owns AI in the C-suite and whether…
igazgatósági tagThe board needs to decide whether AI ownership is a senior executive role or a coordinating function.HR-vezetőBecause of the workforce implications of the AI role, the CHRO should be actively involved in assigning ownership.case · United StatesWho is liable when an AI agent errs? Identity and access management, accountable owner, data breach
Industry sources say AI agents also need a unique identity, an accountable owner and a traceable approval path. Axios warns of unauthorised agents spreading in health care, and ISMG discusses who…
középvezetőThey need to know which agents work in their team and who approved their access.research · globalExecutives use AI more than their staff do, but many overstate their knowledge
According to research reported by CFO.com, senior executives use AI more often than their staff, yet a significant share admit they show more of their knowledge than they really have (5 Oct 2026). HR…
HR-vezetőExecutive AI training should name the knowledge gap openly and give a safe space to admit it.survey · globalThoughtworks: CIOs take on more of the AI workforce redesign as governance lags
A Thoughtworks survey finds that CIOs bear a larger share of responsibility for redesigning the workforce than for core IT infrastructure (Oct. 7, 2026). A separate report says India leads globally…
HR-vezetőAgree a shared accountability and collaboration model with the CIO for workforce redesign.survey · globalMiddle managers almost twice as likely as executives to want humans in the loop on security-sensitive AI decisions
A survey found that middle managers are nearly twice as likely as executives to want a human involved in security-sensitive AI decisions (Oct. 7, 2026).
középvezetőAccording to the survey, it is worth asking for an approval step in sensitive AI decisions and agreeing this with leadership in advance.tool · United StatesMicrosoft unveils Microsoft-Decision-1, a model built for fast decision-making
Microsoft presented Microsoft-Decision-1, a model designed for fast decisions that produces structured output software can use directly (Oct. 9, 2026).
survey · United StatesFew CHROs feel confident they can manage AI-driven change
HR Dive reports that few US chief human resources officers feel very confident managing AI-driven change; according to the excerpt, only 4% of respondents do (Oct. 7, 2026). The main obstacle is the…
HR-vezetőMeasure employee sentiment through direct data collection rather than relying on leaders' impressions.középvezetőAddress team concerns openly and do not read them as excitement.állásfoglalás · globálisAWS executive: companies still don't trust AI with major decisions
Times of India, citing GeekWire, reports that the general consensus at the Madrona IA40 summit was that companies do not yet trust AI with important decisions (Oct. 6, 2026). According to the AWS…