Ethics & Policy

Improving Labor Transparency in AI through Worker Inclusion

Published

2 weeks ago

August 21, 2025

There is widespread public attention to AI’s potential impact on jobs. Everyone is asking questions like, could AI eliminate massive numbers of jobs? How might it change the nature of many jobs? Will it dramatically restructure the labor market? But despite the high level of public interest in these questions, current AI transparency efforts do not cover AI’s labor impacts.

Documentation is one of the only tools that the AI industry & AI research field have widely agreed upon for the assurance of safe and responsible AI systems. But impacts on workers are largely absent from transparency and documentation efforts. In PAI’s latest report, “Workers Participating in Transparency: Addressing the Gap in AI Transparency on Labor,” we explore how to close this gap.

Addressing the Labor Gap in AI Transparency Efforts Panel Discussion

The Importance of Including Workers in AI Transparency Efforts

First, it’s important to understand what we mean by transparency. There are two important concepts of transparency to consider: human-centered transparency and transparency as a process.

The concept of human-centered transparency developed by Liao and Wortman Vaughan starts from the key questions of who and what transparency is for. Transparency is fundamentally for the sake of human understanding, and different stakeholders will have different needs to support their understanding. Critically, Machine Learning (ML) lay users will have different needs than ML experts. Workers are important users and impacted people of ML systems. From a human-centered transparency framework, it’s important to include them and take into account their specific needs.

The concept of transparency as a process, not just an artifact, comes from PAI’s ABOUT ML reference document. Transparency is a process that involves stakeholders in an ongoing critical process of asking, answering and documenting questions about a product and the potential impacts of choices in its design, development and deployment. Workers who are included in this participatory process have key insights to offer as “domain experts” with first-hand experience.

Broadly speaking, worker inclusion can benefit companies. The Ford Foundation, informed by a multisectoral group of corporate leaders, as well as PAI, released a report called Listen to Lead: Raise Retention and Boost Business. The report concludes that engaging workers by listening to them, taking action and being accountable can result in lower turnover, better productivity and more revenue.

Worker involvement in the development and adoption of technology also brings benefits. MIT Sloan School of Management professor Thomas Kochan has a body of research across multiple industries showing that incorporating end users, such as workers, into technological development and deployment results in better products, better implementation, and better jobs. Kochan et al. argue in a recent study that generative AI provides an even bigger opportunity for including worker voice in beneficial ways. As Japanese manufacturers described their philosophy of worker participation in introducing new technology, “it is workers who give wisdom to the machines.”

Including Workers as an Audience and Topic of Transparency

There are three pathways by which we can begin to address the gap in labor transparency – finding ways for workers to be included as a topic of transparency, an audience for transparency and participants in transparency.

One example of including workers as a topic of transparency is the system card for Open AI’s GPT-4. The system card is noteworthy for emphasizing that the impact of GPT-4 on the workforce should be “a crucial consideration” for policymakers and stakeholders. It delves into the potential impact of the model on job automation, job quality and inequality. Open AI also provides transparency about its labor practices for data workers that contribute to the model.

“… engaging workers by listening to them, taking action and being accountable can result in lower turnover, better productivity and more revenue.”

Another example where workers are an audience of and participants in transparency is the model fact sheet for a healthcare tool called Sepsis Watch. Part of Sepsis Watch is a machine learning tool that helps diagnose sepsis, a serious infection that is the leading cause of inpatient death in US hospitals. Healthcare workers helped design the model fact sheet, which resembled a pharmaceutical drug warning label, to convey key information to frontline workers using the tool.

Union Collective Bargaining Offers a Model for Participatory Processes

Collective bargaining is a participatory process by which workers in a union come together as a group and negotiate their working conditions with their employer. Workers help shape the union’s bargaining proposals and ratify the final contract. According to the International Labor Organization, one in three employees in 98 countries are covered under a collective bargaining agreement. As a widespread mechanism, collective bargaining provides an opportunity to address the use of technology in the workplace – see Lisa Kresge’s paper on union bargaining around technology for many examples. For instance, workers have the right to make information requests to the employer during bargaining that could provide greater transparency about technological changes.

Some unions and employers form labor management committees or partnerships that can engage in study or planning around technological change. The Kaiser Labor Management Partnership is one example in the healthcare industry. This union and employer partnership is charged with studying job trends and changing skills required by new technology and promoting job security and workforce training in response to forecasted changes. Such collaboration can also provide a model where workers and employers together can push for greater transparency around labor impacts from model developers.

There is much more we can do to improve labor transparency by including workers as a topic, audience and participants in transparency. Our latest report offers a few examples and explores where further experimentation and research are needed. However, shifting the paradigm and treating transparency as a human-centered process of critical inquiry and documentation can bolster transparency efforts beyond just labor transparency. Taking action to include workers as a topic, audience and participants in transparency ultimately helps us move transparency efforts from a checklist of different artifacts to a process for shared governance that empowers workers.

Source link

Ethics & Policy

5 interesting stats to start your week

Published

3 hours ago

September 1, 2025

Emily Manock

Third of UK marketers have ‘dramatically’ changed AI approach since AI Act

More than a third (37%) of UK marketers say they have ‘dramatically’ changed their approach to AI, since the introduction of the European Union’s AI Act a year ago, according to research by SAP Emarsys.

Additionally, nearly half (44%) of UK marketers say their approach to AI is more ethical than it was this time last year, while 46% report a better understanding of AI ethics, and 48% claim full compliance with the AI Act, which is designed to ensure safe and transparent AI.

The act sets out a phased approach to regulating the technology, classifying models into risk categories and setting up legal, technological, and governance frameworks which will come into place over the next two years.

However, some marketers are sceptical about the legislation, with 28% raising concerns that the AI Act will lead to the end of innovation in marketing.

Source: SAP Emarsys

Shoppers more likely to trust user reviews than influencers

Nearly two-thirds (65%) of UK consumers say they have made a purchase based on online reviews or comments from fellow shoppers, as opposed to 58% who say they have made a purchase thanks to a social media endorsement.

Sports and leisure equipment (63%), decorative homewares (58%), luxury goods (56%), and cultural events (55%) are identified as product categories where consumers are most likely to find peer-to-peer information valuable.

Accurate product information was found to be a key factor in whether a review was positive or negative. Two-thirds (66%) of UK shoppers say that discrepancies between the product they receive and its description are a key reason for leaving negative reviews, whereas 40% of respondents say they have returned an item in the past year because the product details were inaccurate or misleading.

According to research by Akeeno, purchases driven by influencer activity have also declined since 2023, with 50% reporting having made a purchase based on influencer content in 2025 compared to 54% two years ago.

Source: Akeeno

77% of B2B marketing leaders say buyers still rely on their networks

When vetting what brands to work with, 77% of B2B marketing leaders say potential buyers still look at the company’s wider network as well as its own channels.

Given the amount of content professionals are faced with, they are more likely to rely on other professionals they already know and trust, according to research from LinkedIn.

More than two-fifths (43%) of B2B marketers globally say their network is still their primary source for advice at work, ahead of family and friends, search engines, and AI tools.

Additionally, younger professionals surveyed say they are still somewhat sceptical of AI, with three-quarters (75%) of 18- to 24-year-olds saying that even as AI becomes more advanced, there’s still no substitute for the intuition and insights they get from trusted colleagues.

Since professionals are more likely to trust content and advice from peers, marketers are now investing more in creators, employees, and subject matter experts to build trust. As a result, 80% of marketers say trusted creators are now essential to earning credibility with younger buyers.

Source: LinkedIn

Business confidence up 11 points but leaders remain concerned about economy

Business leader confidence has increased slightly from last month, having risen from -72 in July to -61 in August.

The IoD Directors’ Economic Confidence Index, which measures business leader optimism in prospects for the UK economy, is now back to where it was immediately after last year’s Budget.

This improvement comes from several factors, including the rise in investment intentions (up from -27 in July to -8 in August), the rise in headcount expectations from -23 to -4 over the same period, and the increase in revenue expectations from -8 to 12.

Additionally, business leaders’ confidence in their own organisations is also up, standing at 1 in August compared to -9 in July.

Several factors were identified as being of concern for business leaders; these include UK economic conditions at 76%, up from 67% in May, and both employment taxes (remaining at 59%) and business taxes (up to 47%, from 45%) continuing to be of significant concern.

Source: The Institute of Directors

Total volume of alcohol sold in retail down 2.3%

The total volume of alcohol sold in retail has fallen by 2.3% in the first half of 2025 compared to the previous year, equivalent to 90 million fewer litres. Value sales are also down by 1.1% compared to the same period in 2024.

At the same time, retail sales of non-alcoholic drinks have increased by 5.5% compared to last year, while volume sales are up by 2.3%, equivalent to a further 1.5 billion litres.

As the demand for non-alcoholic beverages grows, people increasingly expect these options to be available in their local bars and restaurants, with 55% of Brits and Europeans now expecting bars to always serve non-alcoholic beer.

As well as this, there are shifts happening within the alcoholic beverages category with value sales of no and low-alcohol spirits rising by 16.1%, and sales of ready-to-drink spirits growing by 11.6% compared to last year.

Source: Circana

Source link

Ethics & Policy

AI ethics under scrutiny, young people most exposed

Published

3 hours ago

September 1, 2025

Claire Lemaire

New reports into the rise of artificial intelligence (AI) showed incidents linked to ethical breaches have more than doubled in just two years.

At the same time, entry-level job opportunities have been shrinking, partly due to the spread of this automation.

AI is moving from the margins to the mainstream at extraordinary speed and both workplaces and universities are struggling to keep up.

Tools such as ChatGPT, Gemini and Claude are now being used to draft emails, analyse data, write code, mark essays and even decide who gets a job interview.

Alongside this rapid rollout, a March report from McKinsey, one by the OECD in July and an earlier Rand report warned of a sharp increase in ethical controversies — from cheating scandals in exams to biased recruitment systems and cybersecurity threats — leaving regulators and institutions scrambling to respond.

The McKinsey survey said almost eight in 10 organisations now used AI in at least one business function, up from half in 2022.

While adoption promises faster workflows and lower costs, many companies deploy AI without clear policies. Universities face similar struggles, with students increasingly relying on AI for assignments and exams while academic rules remain inconsistent, it said.

The OECD’s AI Incidents and Hazards Monitor reported that ethical and operational issues involving AI have more than doubled since 2022.

Common concerns included accountability — who is responsible when AI errs; transparency — whether users understand AI decisions; and fairness, whether AI discriminates against certain groups.

Many models operated as “black boxes”, producing results without explanation, making errors hard to detect and correct, it said.

In workplaces, AI is used to screen CVs, rank applicants, and monitor performance. Yet studies show AI trained on historical data can replicate biases, unintentionally favouring certain groups.

Rand reported that AI was also used to manipulate information, influence decisions in sensitive sectors, and conduct cyberattacks.

Meanwhile, 41 per cent of professionals report that AI-driven change is harming their mental health, with younger workers feeling most anxious about job security.

LinkedIn data showed that entry-level roles in the US have fallen by more than 35 per cent since 2023, while 63 per cent of executives expected AI to replace tasks currently done by junior staff.

Aneesh Raman, LinkedIn’s chief economic opportunity officer, described this as “a perfect storm” for new graduates: Hiring freezes, economic uncertainty and AI disruption, as the BBC reported August 26.

LinkedIn forecasts that 70 per cent of jobs will look very different by 2030.

Recent Stanford research confirmed that employment among early-career workers in AI-exposed roles has dropped 13 per cent since generative AI became widespread, while more experienced workers or less AI-exposed roles remained stable.

Companies are adjusting through layoffs rather than pay cuts, squeezing younger workers out, it found.

In Belgium, AI ethics and fairness debates have intensified following a scandal in Flanders’ medical entrance exams.

Investigators caught three candidates using ChatGPT during the test.

Separately, 19 students filed appeals, suspecting others may have used AI unfairly after unusually high pass rates: Some 2,608 of 5,544 participants passed but only 1,741 could enter medical school. The success rate jumped to 47 per cent from 18.9 per cent in 2024, raising concerns about fairness and potential AI misuse.

Flemish education minister Zuhal Demir condemned the incidents, saying students who used AI had “cheated themselves, the university and society”.

Exam commission chair Professor Jan Eggermont noted that the higher pass rate might also reflect easier questions, which were deliberately simplified after the previous year’s exam proved excessively difficult, as well as the record number of participants, rather than AI-assisted cheating alone.

French-speaking universities, in the other part of the country, were not concerned by this scandal, as they still conduct medical entrance exams entirely on paper, something Demir said he was considering going back to.

Source link

Ethics & Policy

Governing AI with inclusion: An Egyptian model for the Global South

Published

5 hours ago

September 1, 2025

Sahar Albazar

When artificial intelligence tools began spreading beyond technical circles and into the hands of everyday users, I saw a real opportunity to understand this profound transformation and harness AI’s potential to benefit Egypt as a state and its citizens. I also had questions: Is AI truly a national priority for Egypt? Do we need a legal framework to regulate it? Does it provide adequate protection for citizens? And is it safe enough for vulnerable groups like women and children?

These questions were not rhetorical. They were the drivers behind my decision to work on a legislative proposal for AI governance. My goal was to craft a national framework rooted in inclusion, dialogue, and development, one that does not simply follow global trends but actively shapes them to serve our society’s interests. The journey Egypt undertook can offer inspiration for other countries navigating the path toward fair and inclusive digital policies.

Egypt’s AI Development Journey

Over the past five years, Egypt has accelerated its commitment to AI as a pillar of its Egypt Vision 2030 for sustainable development. In May 2021, the government launched its first National AI Strategy, focusing on capacity building, integrating AI in the public sector, and fostering international collaboration. A National AI Council was established under the Ministry of Communications and Information Technology (MCIT) to oversee implementation. In January 2025, President Abdel Fattah El-Sisi unveiled the second National AI Strategy (2025–2030), which is built around six pillars: governance, technology, data, infrastructure, ecosystem development, and capacity building.

Since then, the MCIT has launched several initiatives, including training 100,000 young people through the “Our Future is Digital” programme, partnering with UNESCO to assess AI readiness, and integrating AI into health, education, and infrastructure projects. Today, Egypt hosts AI research centres, university departments, and partnerships with global tech companies—positioning itself as a regional innovation hub.

AI-led education reform

AI is not reserved for startups and hospitals. In May 2025, President El-Sisi instructed the government to consider introducing AI as a compulsory subject in pre-university education. In April 2025, I formally submitted a parliamentary request and another to the Deputy Prime Minister, suggesting that the government include AI education as part of a broader vision to prepare future generations, as outlined in Egypt’s initial AI strategy. The political leadership’s support for this proposal highlighted the value of synergy between decision-makers and civil society. The Ministries of Education and Communications are now exploring how to integrate AI concepts, ethics, and basic programming into school curricula.

From dialogue to legislation: My journey in AI policymaking

As Deputy Chair of the Foreign Affairs Committee in Parliament, I believe AI policymaking should not be confined to closed-door discussions. It must include all voices. In shaping Egypt’s AI policy, we brought together:

The private sector, from startups to multinationals, will contribute its views on regulations, data protection, and innovation.
Civil society – to emphasise ethical AI, algorithmic justice, and protection of vulnerable groups.
International organisations, such as the OECD, UNDP, and UNESCO, share global best practices and experiences.
Academic institutions – I co-hosted policy dialogues with the American University in Cairo and the American Chamber of Commerce (AmCham) to discuss governance standards and capacity development.

From recommendations to action: The government listening session

To transform dialogue into real policy, I formally requested the MCIT to host a listening session focused solely on the private sector. Over 70 companies and experts attended, sharing their recommendations directly with government officials.

This marked a key turning point, transitioning the initiative from a parliamentary effort into a participatory, cross-sectoral collaboration.

Drafting the law: Objectives, transparency, and risk-based classification

Based on these consultations, participants developed a legislative proposal grounded in transparency, fairness, and inclusivity. The proposed law includes the following core objectives:

Support education and scientific research in the field of artificial intelligence
Provide specific protection for individuals and groups most vulnerable to the potential risks of AI technologies
Govern AI systems in alignment with Egypt’s international commitments and national legal framework
Enhance Egypt’s position as a regional and international hub for AI innovation, in partnership with development institutions
Support and encourage private sector investment in the field of AI, especially for startups and small enterprises
Promote Egypt’s transition to a digital economy powered by advanced technologies and AI

To operationalise these objectives, the bill includes:

Clear definitions of AI systems
Data protection measures aligned with Egypt’s 2020 Personal Data Protection Law
Mandatory algorithmic fairness, transparency, and auditability
Incentives for innovation, such as AI incubators and R&D centres

Establishment of ethics committees and training programmes for public sector staff

The draft law also introduces a risk-based classification framework, aligning it with global best practices, which categorises AI systems into three tiers:

1. Prohibited AI systems – These are banned outright due to unacceptable risks, including harm to safety, rights, or public order.

2. High-risk AI systems – These require prior approval, detailed documentation, transparency, and ongoing regulatory oversight. Common examples include AI used in healthcare, law enforcement, critical infrastructure, and education.

3. Limited-risk AI systems – These are permitted with minimal safeguards, such as user transparency, labelling of AI-generated content, and optional user consent. Examples include recommendation engines and chatbots.

This classification system ensures proportionality in regulation, protecting the public interest without stifling innovation.

Global recognition: The IPU applauds Egypt’s model

The Inter-Parliamentary Union (IPU), representing over 179 national parliaments, praised Egypt’s AI bill as a model for inclusive AI governance. It highlighted that involving all stakeholders builds public trust in digital policy and reinforces the legitimacy of technology laws.

Key lessons learned

Inclusion builds trust – Multistakeholder participation leads to more practical and sustainable policies.
Political will matters – President El-Sisi’s support elevated AI from a tech topic to a national priority.
Laws evolve through experience – Our draft legislation is designed to be updated as the field develops.
Education is the ultimate infrastructure – Bridging the future digital divide begins in the classroom.
Ethics come first – From the outset, we established values that focus on fairness, transparency, and non-discrimination.

Challenges ahead

As the draft bill progresses into final legislation and implementation, several challenges lie ahead:

Training regulators on AI fundamentals
Equipping public institutions to adopt ethical AI
Reducing the urban-rural digital divide
Ensuring national sovereignty over data
Enhancing Egypt’s global role as a policymaker—not just a policy recipient

Ensuring representation in AI policy

As a female legislator leading this effort, it was important for me to prioritise the representation of women, youth, and marginalised groups in technology policymaking. If AI is built on biased data, it reproduces those biases. That’s why the policymaking table must be round, diverse, and representative.

A vision for the region

I look forward to seeing Egypt:

Advance regional AI policy partnerships across the Middle East and Africa
Embedd AI ethics in all levels of education
Invest in AI for the public good

Because AI should serve people—not control them.

Better laws for a better future

This journey taught me that governing AI requires courage to legislate before all the answers are known—and humility to listen to every voice. Egypt’s experience isn’t just about technology; it’s about building trust and shared ownership. And perhaps that’s the most important infrastructure of all.

The post Governing AI with inclusion: An Egyptian model for the Global South appeared first on OECD.AI.

Source link