What’s happening? Just 2% of companies currently meet Responsible AI (RAI) standards, according to a study by the Infosys Knowledge Institute. The analysis involved a survey of 1,500 executives representing firms with annual revenues above $1bn across 14 industries and seven countries. It found that 95% have experienced AI-related safety, ethical or environmental incidents, while 77% reported financial losses due to poorly implemented AI. Only 38% are taking steps to reduce the environmental impact of AI, revealing a major oversight in ESG strategies. However, companies investing in RAI leadership have seen 39% lower incident costs and reduced risk severity, indicating that ethical and sustainable AI can offer tangible business value alongside compliance. (edie)

Why does this matter? Firms are rapidly adopting AI, 78% use it in at least one function, up from 72% in early 2024 and 55% in 2023, McKinsey research indicates. Generative AI use has surged to 71%, driven by high-value areas such as marketing and sales, service operations, product and service development and software engineering. Companies are also increasingly addressing gen AI risks, prioritising concerns perceived to have the most negative consequences – inaccuracy, cybersecurity and IP infringements. Environmental concerns rank far lower, despite data centres requiring huge amounts of energy and water to remain operational.

RAI as an opportunity – While no universal ethical standard exists, key principles are emerging. These include avoiding bias in datasets, allowing users to understand how algorithms work and preventing harm to individuals, society and the environment. Responsible implementation of AI, system security, data protection and promoting inclusivity in identifying and addressing ethical risks are also highly important. As AI becomes central to business, companies could view RAI as an opportunity rather than a hurdle. Indeed, Infosys’ analysis revealed that 78% of executives believe that RAI is a “critical enabler for business growth.” It also found that firms investing in larger RAI teams are more likely to have a higher number of successful AI deployments and reduced financial losses when AI-related incidents occur.

Regulatory affairs – Companies are also open to new AI regulations, which would provide clarity, trust and confidence inside their business and for their customers. The EU is leading the way, becoming the first jurisdiction to implement a comprehensive AI law through the EU AI Act. The law began taking effect in August 2024, banning high-risk practices such as scraping facial images from public sources by February 2025. As of 2 August 2025, it also covers general-purpose AI models posing systemic risks, including those from OpenAI, Google, Meta and Anthropic. Violations can lead to fines of up to €35m ($40m) or 7% of global turnover. While viewed as a milestone, experts argue that the law gives too little weight to environmental impacts.

Global intentions – International RAI efforts are also underway. Earlier this year, the Paris AI Action Summit saw almost 100 countries and more than a thousand stakeholders to discuss critical issues around AI development, protecting individual freedoms and aligning AI with humanitarian values. The event saw the publication of the International AI safety report, the launch of the Current AI initiative backed with a $400m public-private investment and the formation of an environmental sustainability coalition. The AI Action Summit Declaration, a statement regarding inclusive and sustainable AI, was also endorsed by 60 countries, although the UK and US declined to sign it, citing concerns over national security and global governance. India will host the next AI summit in February 2026, where the country is expected to advocate for a standardised framework for AI deployment in governance and communication.

The agents are coming – AI continues to evolve rapidly, with agentic AI – systems capable of achieving specified goals with minimal oversight – already making inroads. The tech is set to boom in the coming years and will undoubtedly transform the workplace. This brings its own set of challenges which must be considered through an RAI lens. Companies that act now will be best prepared for when agentic AI hits the mainstream.