If people are priced out of neurotechnological products and services, the data collected will exclude them and any future changes made based on this data will likely be biased. As medical intervention with neurotechnology is still developing, reforms and upgrades may be made to the technology predicated on such biased data – favouring those who can afford neurotechnologies and disenfranchising those who cannot. Harvard Researchers have discussed how algorithmic bias will be evident in any form of AI as it impacts medical data. Bias must be acknowledged and mitigated to ensure that it does not exclude, oppress or denigrate vulnerable populations.124 The UK ICO raises the prospect that data will largely be harvested from neurotypical people, leaving neurodivergent customers with potentially biased and ill‑equipped products.125 Discrimination may also take place if devices are not trialled on groups 18 of people who are representative of the general population.126 If algorithmic bias arises, affecting the use of AI in neurotechnologies, it may become embedded in neural devices. This is particularly so where neurotechnologies are used on children, people with neurological disorders or socially marginalised individuals.127 Regulations should be in place to ensure that researchers and companies are actively working to prevent the integration of bias into future products, services or upgrades. This can be assisted by ensuring there is regular discussion about possible biases in data collection, that researchers themselves are from diverse backgrounds, and that there is broad awareness of both potential data biases and how to prevent or mitigate data bias. This will go far in protecting the right to equality and non‑discrimination. Protecting Cognition: Background Paper on Human Rights and Neurotechnology