UK intelligence agencies GCHQ, MI5 and MI6 have proposed weakening ‘burdensome’ surveillance laws that limit their ability to train artificial intelligence (AI) models using extensive sets of personal data.
The agencies want the safeguards governing the use of bulk personal datasets (BPDs) to be reduced.
By relaxing safeguards that protect privacy and prevent misuse of sensitive information, the intelligence agencies believe the AI models can help identify potential terrorists and future informants.
The proposed changes were presented to David Anderson, a senior barrister and member of the House of Lords. Anderson found the proposals would remove the requirement for a judge to approve examination and retention of BPDs. This process would be replaced with a faster self-authorisation process.
Anderson said that AI has been used by intelligence agencies for a many years, with machine learning models already trained with BPDs. The increase in the type and volume of these datasets make machine learning tools useful for the agencies. But Anderson acknowledged that existing safeguards make it difficult for the agencies when applied to publicly available datasets that have ‘little or no reasonable expectation of privacy’.
The data gleaned from video-sharing platforms, academic papers, podcasts, public records, and company information could be grouped into a new category of BPDs, according to Anderson.
Amending the law to establish ‘a less burdensome set of safeguards’ was recommended, especially tailored for this new BPD category. Anderson asserted that the proposed changes will have a relatively minor deregulatory impact, but also suggested preserving a level of ministerial and judicial oversight.
BPDs are frequently utilised by MI5, MI6 and GCHQ. These are sourced from a variety of closed and open sources, often obtained through covert means.
Privacy experts and civil liberties groups were alarmed by this move, arguing that it could unwind some legal protections introduced in 2016 following Edward Snowden’s revelation about intrusive state surveillance.
“Data scientists’ disappointment they don’t get to play with all their wonderful new toys is not a good justification for weakening fundamental rights protection,” said privacy and surveillance expert Ian Brown.
Brown criticised the amount of time officials had dedicated to ‘directly delivering BPD authorisations’ to facilitate the intelligence agencies’ access to new datasets.
Lord Anderson’s recommendations are being reviewed by the UK Government, with a response expected later this year.
The UK government recently introduced its pro-innovation White Paper on AI Regulation. This White Paper sparked mixed reactions from various stakeholders.
Hungry for more tech news?
Sign up for your weekly tech briefings!
