Extended Intelligence: A new framework for human-AI collaboration
Artificial Intelligence has been a prevalent topic in recent years. Whilst progress in AI technology has accelerated markedly, discourse surrounding AI often adopts an unhelpful oppositional framing, positioning it as rival to human cognitive capabilities.
A growing number of researchers and technologists are challenging this view. The concept of Extended Intelligence (EI) offers a simpler way of thinking about what AI is and how it should be used. It is also a framework that has an impact on how organisations, legal teams and policymakers approach AI responsibly.
There is a difference between "Artificial" Intelligence and "Extended" Intelligence. "Artificial" suggests something synthetic and separate from human thought. "Extended" suggests the opposite, in that AI is a continuation of what we can do and not a challenger to it. The case for Extended Intelligence is simply that AI should be seen in the same light, as a tool that helps us to do more, not one that replaces human intelligence.
At its core, EI is about augmentation over replacement. The goal is to amplify human reasoning, creativity, problem-solving and decision-making without reducing human agency. Just as a calculator extends our arithmetic abilities or a telescope extends our vision, AI should be understood as an extension of the human mind rather than a substitute for it.
When AI is built with this in mind, the human stays in control. The system is there to inform and assist, not to decide. That matters a great deal when it comes to questions of accountability and trust.
A collaborative model
Extended Intelligence treats the relationship between humans and AI as one of partnership rather than competition. AI is good at handling large amounts of data, spotting patterns and carrying out repetitive tasks quickly. Humans are good at making judgements, thinking creatively, understanding context and weighing up ethical considerations. Together, these strengths complement each other well.
This combination is sometimes described as a human-AI hybrid intelligence. The idea is that by letting AI handle data-intensive or repetitive work, humans are freed up to focus on higher-level thinking, strategy, ethics and innovation. Neither party is redundant, instead both are working to their strengths.
In legal practice, for example, AI can help identify relevant cases, spot inconsistencies or assess the likely implications of different contractual terms. The lawyer then brings their professional knowledge, understanding of the client and strategic thinking to bear. Neither replaces the other; both do their job better as a result of working together.
Where the idea comes from
The thinking behind Extended Intelligence goes back further than the current AI boom. In the 1960s, J.C.R. Licklider wrote about the idea of humans and computers working closely together, each making up for what the other lacked. Douglas Engelbart, who invented the computer mouse, built his entire career around the idea that computers should extend the reach of the human mind rather than operate independently of it.
These were the original advocates of what we now call Extended Intelligence. Their focus was on useful tools in human hands, not on autonomous machines. That same thinking can be seen in the work of some of today's AI developers. Organisations such as xAI (the company behind Grok), have been clear that their goal is to use AI to advance human understanding and scientific discovery, not to build systems that work independently of human direction. xAI's mission positions AI as a means of understanding the universe by extending human capability, a deliberate contrast to the pursuit of autonomous superintelligence.
What this means in practice
For organisations thinking about how to adopt AI responsibly, Extended Intelligence provides a practical framework to guide those decisions.
Transparency is the first consideration. If AI is there to support human judgement, the people using it need to be able to understand how it reaches its conclusions. The EU AI Act already requires that high-risk AI applications be explainable, and such regulatory standards are likely to become more demanding over time.
Oversight is the second. Keeping humans involved in decisions is becoming a legal consideration. Regulators across financial services, healthcare and legal fields are increasingly focused on whether significant decisions are subject to meaningful human review rather than these just being left to automated processes.
Purpose is the third. Extended Intelligence invites a straightforward question at every stage of AI adoption: does this tool actually help the people using it? This question should inform which tools an organisation chooses, how it governs their use and how it explains that use to clients and stakeholders.
There is also a broader point worth making, as much of the public conversation about AI is still driven by fear of job loss. Extended Intelligence offers a more grounded alternative. Rather than science-fiction scenarios of AI takeover, EI focuses on practical benefits: accelerating research in medicine, physics and climate science through AI-assisted tools and helping organisations to make better decisions faster.
Law is a profession built on judgement, precedent and professional responsibility. Therefore, these things will not simply be taken over by AI. However, the amount of legal information that practitioners are expected to navigate is growing and the pressure to work efficiently has not eased. Extended Intelligence provides legal teams with a sensible way to consider AI. It is clear-eyed about what it can or cannot do and always keeps the lawyer and the client's interests at the centre. That is more useful than either dismissing AI entirely or adopting it without proper thought.
Ultimately, Extended Intelligence is not just a technical term for specialists, but a straightforward position on the relationship between people and the tools they use, as well as what we want AI to look like as it becomes a bigger part of working life.
EI reframes AI as a seamless extension of our own intelligence, promoting a future where humans and AI develop together to achieve greater collective understanding and progress. This is particularly relevant to discussions about responsible AI deployment, and to ensuring that technology serves humanity's goals rather than operating outside of them.
For anyone navigating AI adoption, whether in business, law or policy, it offers a clear set of priorities. The most powerful AI is not the most independent - it is the kind that works best in the hands of the people using it.