On 5 October 2017, the Coalition government (under Prime Minister Malcolm Turnbull) did approve a national facial recognition system through the Council of Australian Governments (COAG) [1].
The system, formally called the "National Facial Biometric Matching Capability," was established via the Intergovernmental Agreement on Identity Matching Services [2].
**Did it collate CCTV footage?** This is partially misleading.
The primary data sources for the facial recognition system were driver's license photographs and passport images from state and territory governments, not CCTV cameras [3].
Turnbull explicitly mentioned CCTV in his comments, saying the system would "work in concert with CCTV footage" [1], but the actual agreement focused on government-held identity databases, not broad CCTV integration.
**Did it share data with private companies?** The agreement explicitly stated that "authorised private sector organisations" could be given access to face verification services for document verification purposes [1].
This was a stated feature of the system, conditional on government approval and with "consent of the individual concerned" (though critics disputed the adequacy of these safeguards) [1].
**The "doesn't involve surveillance" claim.** Turnbull did make this exact statement.
This statement reflected the government's position that the system was merely automating existing inter-agency identity verification procedures rather than creating new surveillance capabilities [1].
However, the claim omits critical context that substantially affects its interpretation:
**On the "surveillance" terminology:** The government's claim about "not involving surveillance" requires context.
The system was fundamentally about matching faces in government identity databases against each other to verify identity—a database-matching function distinct from CCTV surveillance in the sense of monitoring public spaces [3].
The ambiguity lies in whether "surveillance" refers to the collection and tracking enabled by such data consolidation (critics' view) or narrowly to the act of monitoring people in public (government's view) [1][4].
**On the origins and scope:** The agreement was signed at a special COAG meeting focused on counter-terrorism and was presented as part of an anti-terrorism framework, not as a broad surveillance expansion [2].
However, the agreement explicitly allowed the system to be used for law enforcement generally, not just terrorism [1].
**On privacy safeguards:** The government framed the system as including "robust privacy and security safeguards" [2], though critics immediately disputed whether these safeguards were adequate.
The parliamentary joint committee on intelligence and security and Labor's shadow attorney-general both flagged concerns and requested detailed examination [1].
**On the "increases privacy" claim:** The government's argument was that preventing identity crime (via biometric matching) would protect individual privacy by preventing identity theft [2].
This is a distinct claim from "surveillance increases privacy"—it argues identity-matching prevents privacy violations, not that surveillance itself protects privacy.
The original sources provided with the claim are both credible but reflect different perspectives:
**The Guardian** [1] is a mainstream, respected news organization with strong reputation for accuracy.
The Guardian Australia has editorial perspective on civil liberties issues but the reporting here is factual.
**Techdirt** [4] is a specialized technology and civil liberties blog founded by Mike Masnick.
While tech-savvy and articulate on privacy issues, Techdirt explicitly expresses strong editorial perspective against government surveillance programs.
The article is titled "Australian Government Claims That Facial Recognition Systems Increase Privacy..." with the ellipsis conveying skepticism, and the commentary section contains strong criticism.
**Did Labor establish facial recognition systems?**
Search conducted: "Labor government facial recognition surveillance Australia policy"
Labor has NOT established a comparable facial recognition system during their time in opposition (2013-2022).
* * * *
However, this is not a meaningful comparison because Labor was in opposition when the 2017 agreement was signed—they had no opportunity to implement it [1].
More relevant is Labor's **record on surveillance policies when in government (2007-2013)**: Labor approved various data-sharing and identity security measures, though not a facial recognition system specifically.
When Turnbull's proposal came before Parliament, Labor's shadow attorney-general Mark Dreyfus stated the measures "appear sensible" but wanted to see detail, and did not commit to opposing the system [1].
**Contemporary Labor position:** Labor did not vocally oppose the 2017 facial recognition agreement on principle, though they requested parliamentary scrutiny.
The claim requires significant unpacking:
**The criticism is justified but overstated.** Critics are correct that:
- The government introduced a facial recognition system that consolidated biometric data [1][2]
- Private sector access was explicitly permitted [1]
- Turnbull's claim that this "doesn't involve surveillance" is either misleading or reflects a narrow definition of surveillance [1][4]
- The government's claim that identity-matching "increases privacy" is philosophically questionable [4]
However, the claim conflates several distinct issues:
**What the system actually was:** The National Facial Biometric Matching Capability was fundamentally a database-matching system linking existing government identity records (driver's licenses, passports) for automated verification purposes [3].
This is functionally different from:
- Real-time CCTV monitoring (which was not established—facial matching WITH CCTV was only mentioned as a future possibility)
- Warrantless mass surveillance (the system required authorization for access, though safeguards were disputed)
**Turnbull's misleading claim:** Turnbull's statement that the system "doesn't involve surveillance" is defensible only under a narrow technical definition (database matching ≠ surveillance of public spaces) but indefensible under a broader privacy interpretation [1][4].
His framing was clearly misleading or at minimum incomplete.
**The "increases privacy" argument:** The government's logic was: "Preventing identity theft via biometric matching protects individual privacy by preventing criminals from stealing identities" [2].
While critics found this reasoning flawed (because the system enables government and corporate access to biometric data, creating different privacy risks), it's not demonstrably false—it reflects a different conception of privacy [4].
Privacy can meaningfully mean both "protection from identity theft" AND "freedom from surveillance," and the system improves one while potentially worsening the other.
**Comparative context:** Other democratic nations, including the UK, US, and Canada, have implemented or are implementing facial recognition systems [3].
Australia's system was not uniquely aggressive by international standards, though concerns about Australian privacy frameworks are legitimate [4].
**What changed:** The substantive policy change was not that facial recognition existed (it already did) or that government agencies shared data (they already did), but that the process was automated and private sector access was permitted.
The claim accurately identifies that the Coalition government established a national facial recognition system in 2017 with private sector access provisions, and that Turnbull made the specific claim that it "doesn't involve surveillance." These facts are verifiable [1][2].
However, the claim is incomplete and somewhat misleading in key respects: (1) The system primarily used government identity databases rather than "collating faces from CCTV cameras" [3]; (2) The "doesn't involve surveillance" claim, while literally made, reflected a narrow technical definition rather than the broader privacy implications critics emphasized [1][4]; (3) The "will increase citizen's privacy" claim described government reasoning about identity theft prevention rather than a claims that surveillance increases privacy generally [2].
The claim captures a genuine policy concern—government expansion of facial recognition and biometric data consolidation—but overstates the CCTV component and mischaracterizes the government's privacy argument in a way that misses the actual substantive debate.
The claim accurately identifies that the Coalition government established a national facial recognition system in 2017 with private sector access provisions, and that Turnbull made the specific claim that it "doesn't involve surveillance." These facts are verifiable [1][2].
However, the claim is incomplete and somewhat misleading in key respects: (1) The system primarily used government identity databases rather than "collating faces from CCTV cameras" [3]; (2) The "doesn't involve surveillance" claim, while literally made, reflected a narrow technical definition rather than the broader privacy implications critics emphasized [1][4]; (3) The "will increase citizen's privacy" claim described government reasoning about identity theft prevention rather than a claims that surveillance increases privacy generally [2].
The claim captures a genuine policy concern—government expansion of facial recognition and biometric data consolidation—but overstates the CCTV component and mischaracterizes the government's privacy argument in a way that misses the actual substantive debate.