Met investigates hundreds of officers after using Palantir AI tool
The Guardian
Overall Assessment
The article centers on the Met's use of Palantir's AI to identify misconduct, presenting official data and framing the tool as a transformative internal accountability mechanism. It includes some contextual criticism of Palantir but om lacks independent voices or technical details about the AI’s reliability. The tone leans slightly toward institutional justification, with minimal exploration of privacy or due process concerns.
"Palantir has connections to ICE, Donald Trump’s immigration enforcement programme, and to the Israeli military and earlier this month MPs demanded a £330m contract between Palantir and the NHS be scrapped."
Cherry-Picking
Headline & Lead 85/100
The article reports on the Metropolitan Police's use of Palantir's AI software to detect misconduct among officers, resulting in investigations and arrests. It includes official statements and specific figures on violations, while also noting controversies around Palantir's broader ties. The reporting is largely factual but emphasizes the AI's role and the scale of investigations, with limited external critique or alternative perspectives.
✕ Sensationalism: The headline uses 'hundreds of officers' and 'Palantir AI tool' in a way that emphasizes scale and technological novelty, potentially exaggerating the drama of internal police investigations.
"Met investigates hundreds of officers after using Palantir AI tool"
✕ Framing by Emphasis: The lead emphasizes the AI tool as the central actor in uncovering misconduct, potentially overstating its role relative to human oversight and existing procedures.
"The software was deployed by the Met over the course of a week, snooping on staff members using data the force has ready access to, unearthing rule-breaking ranging from work-from-home violations to suspected corruption and even criminal allegations such as rape."
Language & Tone 78/100
The article reports on the Metropolitan Police's use of Palantir's AI software to detect misconduct among officers, resulting in investigations and arrests. It includes official statements and specific figures on violations, while also noting controversies around Palantir's broader ties. The reporting is largely factual but emphasizes the AI's role and the scale of investigations, with limited external critique or alternative perspectives.
✕ Loaded Language: The term 'snooping' implies invasive or unethical surveillance, introducing a negative connotation toward the police's self-auditing actions.
"snooping on staff members using data the force has ready access to"
✕ Editorializing: Describing Palantir as 'controversial' without immediate elaboration inserts a judgmental tone that shapes reader perception before context is given.
"the controversial tech company Palantir"
✓ Balanced Reporting: The article includes direct quotes from Met commissioner Mark Rowley defending the use of AI as necessary for integrity and public trust, providing an official justification.
"“Criminals are constantly adapting how they use technology and policing has to keep pace, not just on the streets but within our own organisation.”"
Balance 70/100
The article reports on the Metropolitan Police's use of Palantir's AI software to detect misconduct among officers, resulting in investigations and arrests. It includes official statements and specific figures on violations, while also noting controversies around Palantir's broader ties. The reporting is largely factual but emphasizes the AI's role and the scale of investigations, with limited external critique or alternative perspectives.
✓ Proper Attribution: Key claims about the number of officers investigated and types of misconduct are directly attributed to the Metropolitan Police.
"According to numbers cited by the Met, corruption was the most consistent offence detected by the AI software"
✓ Comprehensive Sourcing: The article cites both internal police data and includes a direct quote from the Met commissioner, offering authoritative sourcing from the institution conducting the investigations.
"The Met commissioner, Mark Rowley, said: “Criminals are constantly adapting how they use technology and policing has to keep pace, not just on the streets but within our own organisation.”"
✕ Omission: The article does not include perspectives from police unions, civil liberties groups, or independent data ethics experts who might question the use of AI for internal surveillance.
Completeness 65/100
The article reports on the Metropolitan Police's use of Palantir's AI software to detect misconduct among officers, resulting in investigations and arrests. It includes official statements and specific figures on violations, while also noting controversies around Palantir's broader ties. The reporting is largely factual but emphasizes the AI's role and the scale of investigations, with limited external critique or alternative perspectives.
✕ Cherry-Picking: The article highlights Palantir’s connections to ICE and the Israeli military and NHS controversy, but only in a negative light, without discussing potential benefits or neutral assessments of its technology in public sector use.
"Palantir has connections to ICE, Donald Trump’s immigration enforcement programme, and to the Israeli military and earlier this month MPs demanded a £330m contract between Palantir and the NHS be scrapped."
✕ Omission: There is no discussion of how the AI tool works, its error rate, potential for false positives, or oversight mechanisms—key context for evaluating the fairness of the investigations.
✕ Selective Coverage: The article focuses on the novelty of AI use in policing but does not compare it to prior internal investigation methods, which would help assess whether this represents a significant shift or incremental change.
AI is portrayed as highly effective in uncovering police misconduct
[framing_by_emphasis] The AI tool is positioned as the central actor in discovering widespread rule-breaking, with minimal discussion of limitations or risks.
"The software was deployed by the Met over the course of a week, snooping on staff members using data the force has ready access to, unearthing rule-breaking ranging from work-from-home violations to suspected corruption and even criminal allegations such as rape."
Palantir is framed as ethically questionable due to its political and military associations
[cherry_picking] Negative affiliations of Palantir are selectively highlighted without balanced assessment of its public-sector applications.
"Palantir has connections to ICE, Donald Trump’s immigration enforcement programme, and to the Israeli military and earlier this month MPs demanded a £330m contract between Palantir and the NHS be scrapped."
Police force is framed as institutionally compromised, requiring AI-driven internal cleansing
[framing_by_emphasis] The scale of officer investigations is emphasized, suggesting systemic corruption rather than isolated incidents.
"The Metropolitan police have launched investigations into hundreds of officers after using an AI tool built by the controversial tech company Palantir to root out rogue cops."
Trump administration is indirectly framed as adversarial through Palantir's ties to ICE and controversial policies
[cherry_picking] Palantir’s association with Trump-era immigration enforcement is highlighted to imply ethical contamination.
"Palantir has connections to ICE, Donald Trump’s immigration enforcement programme, and to the Israeli military and earlier this month MPs demanded a £330m contract between Palantir and the NHS be scrapped."
Officers under investigation are framed as excluded from procedural protections, with no mention of due process
[omission] The article omits any discussion of oversight mechanisms, error rates, or rights of accused officers, implying guilt-by-detection.
The article centers on the Met's use of Palantir's AI to identify misconduct, presenting official data and framing the tool as a transformative internal accountability mechanism. It includes some contextual criticism of Palantir but om lacks independent voices or technical details about the AI’s reliability. The tone leans slightly toward institutional justification, with minimal exploration of privacy or due process concerns.
The Metropolitan Police have used AI software developed by Palantir to analyze internal data and identify potential misconduct among officers. Investigations have been launched into hundreds of staff members for violations ranging from IT misuse to serious criminal allegations, with three arrests made. The Met stated the tool helps enforce standards, while the article notes Palantir's controversial external partnerships.
The Guardian — Other - Crime
Based on the last 60 days of articles
The Guardian
