Sara Alsherif سارة الشريف
We @openrightsgroup . are publishing today the legal opinion about the Home Office's use of ChatGPT4 in the asylum system decision-making process. The legal opinion concludes that these tools are likely unlawful, which could open the way for a legal challenge.
https://www.independent.co.uk/news/uk/home-news/ai-artificial-intelligence-asylum-claims-backlog-b2937111.html
Home Office use of AI in asylum cases could be unlawful, legal experts warn

Exclusive: Caseworkers use AI to summarise asylum interviews and to search for information about country of origin

The Independent
Mar 16 at 2:33pmWeb
Show threadSara Alsherif سارة الشريف3d ago
The legal opinion analysis leads to the conclusion that, in significant respects, the UK AI Playbook is not being followed, or that there are serious concerns that it is not. Ex, there has been a failure to assess quantitatively the extent to which the Asylum policy search tool (APS) produces inaccurate outputs.
Show threadSara Alsherif سارة الشريف3d ago
The opinion found that in the Asylum case summarisation (ACS), there is insufficient information about what accuracy means in this context, the extent of inaccuracy, and the benchmarking the Government is using, as well as the absence of the ability to cross-reference the summarised output from the ACS and APS.
Show threadSara Alsherif سارة الشريف3d ago
And the worst of all is that the asylum applicants will not be informed that AI tools have been used to summarise their interviews or in the asylum process. So they have no opportunity to see the text generated about them! This is considered an important missing procedural safeguard.
Show threadSara Alsherif سارة الشريف3d ago

You can access the executive summary and read the full legal opinion from this link:

https://www.openrightsgroup.org/publications/legal-opinion-the-use-of-artificial-intelligence-tools-in-asylum-cases/

Legal Opinion: The use of Artificial Intelligence tools in asylum cases

Open Rights Group has commissioned this legal Opinion to explain how civil society actors can evaluate the use of Artificial Intelligence (AI) tools and systems by the Government, whether through dialogue or litigation.

Open Rights Group
Show threadPhil M0OFX10h ago
@sarahalsherif @openrightsgroup Given how bad a job ChatGPT does on anything more complex than "proofread this please", it's good to see someone kicking up a stink about the Home Office of all people using it!