Safer City — Botswana's Nationwide AI-Enabled Police Camera Network

The Safer City project, first flagged in President Ian Khama's 2016 State of the Nation Address, is a police-run HD CCTV and automatic number-plate recognition (ANPR) network deployed by the Botswana Police Service (BPS) with technology vendor Huawei Technologies Botswana. Xinhua reported the system became operational in Gaborone and Francistown as of February 2020, feeding a command centre at Kutlwano Police Station over fibre-optic links.

A BPS spokesperson cited goals of securing life and property and countering terrorism.

Sunday Standard's Botswana-based investigative reporting puts the network at 650 operational cameras in Gaborone and 524 in Francistown, with BPS plans to expand to 2,000 and 974 respectively, covering major roads, intersections and traffic lights. Business Weekly's reporting, citing tender no. DJS/MMTC/PPOL:083/2014-2015, put Phase 1 (Gaborone, roughly a 50km radius) at approximately P3 billion, with the full three-phase national rollout costed at roughly P10 billion; a separate P77.7 million FY2020/21 BPS budget line covered further cameras, LTE links and fingerprint systems.

No crime-reduction statistics, arrest counts, or facial-recognition accuracy figures have been published for Safer City.

Governance of the project has been genuinely contested, not a clean success. On 15 July 2021 Botswana's own procurement regulator, the PPADB, rejected a selective, direct-appointment tender to Huawei Technologies Botswana and ICT Dynamix, ruling the process improper and its scope undefined. The Association for Progressive Communications separately documents Huawei-partnered cameras with facial-recognition capability in use without clear consent or transparency mechanisms, operating in the absence of a dedicated legal framework beyond Botswana's 2018 Data Protection Act; University of Botswana law professor Tachilisa Balule is quoted stating that no law specifically authorises the project. This is a well-documented case of a large-scale, AI-enabled surveillance rollout with real deployment and cost figures, but an unevaluated, legally contested implementation and no independent evidence of public-safety impact.