South Africa governs artificial intelligence with policy, but its intellectual-property law is the most AI-accommodating in the founder set. For a jurisdiction whose AI-specific rules are policy, the IP posture and the international footprint are the profile's centre of gravity.
The shape is a substantial jurisdiction whose general AI law is still only a policy framework, over a binding tier of general law. The distinctive weight is in the intellectual-property statutes, which accommodate machine-made output more than any peer, and in the international band.
South Africa's data-protection statute and the general law most likely to reach an AI harm. Its section 71 gives a data subject a right not to be subject to a decision with legal or substantial effect based solely on automated processing intended to profile them, subject to carve-outs. It is a decision right, not an AI-specific labeling or transparency duty.
Source: gov.zaThe computer-crime law, criminalizing data offences and the distribution of harmful and intimate-image data. Technology-general, with no AI or deepfake-specific provision.
Source: gov.zaThe e-commerce and data-message baseline, the enabling layer for electronic dealings, including AI-mediated ones.
Source: gov.zaThe content law reaching harmful and non-consensual intimate content online. Adjacent to synthetic media but not an AI-provenance or deepfake-marking duty.
Source: gov.zaThe authorship framework, and unusually accommodating. Its definition of author includes, for a computer-generated work, the person by whom the arrangements necessary for the creation of the work are undertaken, a provision inserted in 1992. So machine-generated output has an author and a route to protection, though the rule predates generative AI and no court has applied it to it.
Source: WIPO LexThe patent framework, administered by a non-examining office that checks formalities rather than substance. On that basis CIPC granted patent ZA 2021/03242 in 2021 naming an artificial intelligence, DABUS, as inventor, the only such grant in the world. The same applications were refused on the merits elsewhere, so the grant reflects no substantive endorsement of AI inventorship.
Source: gov.za; CIPCThe flagship AI instrument, a first step toward a full national AI policy. It prioritizes human-centred AI, alignment with the data-protection law, risk-tiering and capacity-building, and it tasks and guides rather than binds.
Source: DCDTThe standing digital-policy reference in the absence of a binding AI strategy, recommending institutions and priorities for AI and related technologies. Advisory.
Source: gov.zaThe long-running copyright reform. It would add a fair-use exception capacious enough to shelter some text-and-data mining, though it carries no dedicated mining exception. It is not in force, so no fair-use or mining lane yet exists.
Source: Constitutional Court; the PresidencyReported founding role; South Africa listed among the twenty-nine founders in Chinese releases, but no South African government confirmation or named signatory was found. Official source: Chinese and allied media roster.
Adopted 22 to 23 Nov 2025 under South Africa's G20 presidency; established a task force on AI, data governance and innovation and welcomed an AI-for-Africa initiative. Official source: g20.org.za; the Johannesburg declaration.
Adopted 7 Jul 2025; South Africa a BRICS member. Official source: brics records; the statement text.
Adopted by the AU Executive Council July 2024; South Africa an AU member. Official source: au.int.
Adopted 2021, as a UNESCO member; readiness assessment completed. Official source: unesco.org.
Adopted by consensus 22 Sep 2024; applies to South Africa as a UN member. Official source: A/RES/79/1 annex, un.org.
Adopted by consensus in 2024; South Africa bound as a UN member. Official source: UN records.
Axes: specificity (x, 0-100) is how squarely the instrument is written for AI; force (y, 0-100) is how hard it binds. Placements are editorial judgments for the comparison plot; the tier column is the authoritative classification. The pattern is a country whose AI-specific instruments are a policy framework and a pending bill low on force, above a binding tier of general law, with two intellectual-property statutes that quietly accommodate machine-made output.
Tap a marker
The binding and prescriptive corner is the crowded one: the CAC-led measures put force and detail together, service by service. The sector guidance sits just left of the binding line, nonbinding in form but treated as obligatory. Dashed markers are drafts; the National AI Law would consolidate the whole stack into one statute.
South Africa has real courts and a searchable body of published judgments, and a small AI docket has already formed. Two adjudicated matters are recorded below, both sanctions for AI-fabricated legal citations, each with a published neutral citation. No copyright, patent or deepfake case turning on AI was found. Per the fabrication screen, a citation is given only where it is published and verified.
Portals and sweeps: the Southern African Legal Information Institute for judgments; WIPO Lex and the patent office for the IP filings; South African legal media of record. South Africa publishes judgments widely, so the negative findings are reasonably firm.
South Africa is the most AI-accommodating jurisdiction in the founder set: its patent office granted a patent naming an AI inventor and its copyright law protects computer-generated works, though both accommodations carry caveats and there is no training exception in force.
The AI-IP Index reads a country on two questions that usually get folded together. Accommodation asks how far the law will go to protect intellectual property that a machine helped produce. Institution asks how much administrative machinery sits behind that law: its guidance, its international submissions, its disclosure duties.
A country can be generous on paper and have built almost nothing, or cautious and highly organised, so the two are scored apart and reported as a pair. Each rests on three questions, scored 0, 0.5 or 1.
How far the law will protect intellectual property that a machine helped make.
How much administrative machinery sits behind that law.
The scale runs the same way on both axes: 0 is the open or developed end, 1 is the restrictive or absent one, and 0.5 sits between for anything conditional or untested. Read as a pair, the two scores show whether AI-assisted work can be protected here, and whether the office has said so in writing.
Within each axis the three questions carry equal weight. That is a deliberate choice, not an oversight. An AI developer might reasonably value freedom to train above patent inventorship, but weighting is where indices lose their credibility, so the score stays a flat mean and the choice is stated in the open. Each cell also cites a statute section, a case or an office page, so any single score can be challenged without disputing the rest.
Two positions are worth naming. A country can be open on the law but thin on machinery, an open-but-undeveloped stance that bets on ambiguity. Or it can be restrictive on rights yet run an active office, a restrictive-but-developed stance that reflects a settled policy choice rather than a gap. The pair of scores tells those two apart where a single ranking would blur them.
A rubric is only as good as the results it produces on cases where the answer is already known. These three jurisdictions sit outside this series and are scored with the same six questions, as a check. Their law is the best documented anywhere, so if the scale puts them where the law actually sits, the scale is doing its job. They are reference points, not editions.
Firm judicial settlement at both poles and the weakest statutory position on training. The courts have closed the door on AI inventors and AI authors, while everything on training rides on unresolved fair-use litigation. There is no per se duty to declare AI use in a filing, only a candour duty that bites where the use is material.
Thaler v. Vidal, 43 F.4th 1207 (Fed. Cir. 2022); Thaler v. Perlmutter (D.C. Cir. 2025), cert. denied 2 Mar 2026; Bartz v. Anthropic (N.D. Cal. 2025); USPTO revised inventorship guidance, 90 Fed. Reg. 54636 (28 Nov 2025); 89 Fed. Reg. 58128 (17 Jul 2024); 89 Fed. Reg. 25609 (11 Apr 2024).
The most permissive statutory position on training and the strongest transparency duty in force, sitting over an authorship question no court has answered. Read the label carefully: training and transparency run through EU law, but patents run through the EPO under the European Patent Convention, which is not an EU institution and covers 39 states.
Directive (EU) 2019/790, arts. 3 and 4; Regulation (EU) 2024/1689, art. 53(1)(c) and (d); EPO J 8/20 (2021); EPO Guidelines G-II 3.3.1 (2026 edition); Infopaq C-5/08; Painer C-145/10.
Liberal on patents since the Supreme Court abandoned the Aerotel test in February 2026, conservative on copyright, and currently thin on AI-specific machinery after the dedicated AI examination guidance was withdrawn. Section 9(3) nominally protects a work with no human author for 50 years, the widest AI-output right on paper, but no court has ever applied it to generative AI.
CDPA 1988 ss. 9(3), 12(7), 29A, 178; Thaler v Comptroller-General [2023] UKSC 49; Comptroller-General v Emotional Perception AI [2026] UKSC 3; UKIPO practice notice, 14 Jul 2026; Data (Use and Access) Act 2025 ss. 135 to 137.
The check earned its keep twice. The untested rule held: section 9(3) of the UK Copyright Act reads like the widest AI-output right anywhere, yet because no court has applied it to generative AI it scores as untested rather than open, which is the correct answer and the one a headline reading would miss. The check also exposed a limit worth stating. The Institution axis measures AI-specific machinery, not general office capacity, so a long-established office can score poorly when it has withdrawn its AI guidance and imposes no disclosure duty. Read that axis as what an office has built for AI, nothing wider.
Each row below carries the test we applied and the finding behind the score, with a source. The rationale, not the number, is the point.
Scores verified 21 July 2026, against AIP Index codebook v1.0. A score is only true as of its date.
What we look forA granted patent or a clearly workable examination route, not merely a silent statute.
The findingSouth Africa is the one jurisdiction where an artificial intelligence has been named as inventor on a granted patent: the patent office granted the DABUS patent in 2021. That is the open pole in practice. The heavy caveat, recorded here rather than in the score, is that the office examines only formalities, so the grant carries no substantive endorsement and the same applications were refused on the merits elsewhere.
What we look forProtection applied where human choices shape the work, and refusal of purely machine output.
The findingThe copyright law protects computer-generated works and assigns authorship to the person by whom the arrangements necessary for their creation are undertaken, so machine-generated output has a route to protection, unlike in the human-author-only jurisdictions. The provision predates generative AI and no court has applied it to a generative-AI output, so it earns the middle rather than the open pole.
What we look forA statutory or judicial mining lane. Silence scores against the miner, since the reserved-rights default governs.
The findingThere is no training or text-and-data-mining exception in force. The fair-use reform that could shelter some mining sits in the Copyright Amendment Bill, which the Constitutional Court upheld in part in June 2026 but which is not yet signed into law, and even that bill carries no dedicated mining exception. With copyright silent, the reserved-rights default governs the use of works for training.
What we look forA traceable, substantive submission on the record, not bare membership.
The findingSouth Africa is a WIPO member and broadly active, but the WIPO Conversation submissions record shows no traceable substantive submission from South Africa or the patent office, and no distinct session participation was found. Membership without a submission earns the middle.
What we look forA standalone AI examination text, not AI handled quietly under general practice.
The findingNo dedicated guidance from the patent office on examining AI-related filings was found, consistent with a non-examining office that checks formalities rather than substance. A reasonable-search negative.
What we look forA mandatory declaration inside the filing, not a proposal or an informal request.
The findingNo filing requires declaring AI use, and no in-force instrument imposes a labeling or transparency duty on AI-generated content. The data-protection law's automated-decision right is not a disclosure duty, and the AI framework is policy.
Method. Six questions, three per axis, scored 0, 0.5 or 1 against AIP Index codebook v1.0. The three questions in each axis are weighted equally by choice, and the two axes are reported as a pair, never blended into one number. Every score cites a statute, case or office page, carries a verified date, and records its direction of travel apart from the level. Positions are editorial judgements read from primary law and office practice, not official scores. WIPO's public catalogue of IP-office AI initiatives does not score jurisdictions on this basis, so the readings here come straight from the source law. Not legal advice.
South Africa governs artificial intelligence with policy, but its intellectual-property law is the most AI-accommodating in the founder set. There is no binding AI statute: a National AI Policy Framework of October 2024 sets direction, and the data-protection law reaches automated decisions, but neither imposes an AI-specific operator duty. What makes South Africa distinctive sits in the patent and copyright system. Its patent office granted the DABUS patent in 2021, naming an artificial intelligence as inventor, the only such grant anywhere, and its copyright law has protected computer-generated works since 1992, assigning authorship to the person who arranged their creation. Both accommodations carry heavy caveats, the patent office examines only formalities and the computer-generated-works rule predates generative AI, but together they place South Africa nearer the open pole than any peer. The country also carries real weight abroad, holding the 2025 G20 presidency whose Johannesburg declaration launched an AI-for-Africa initiative, and it has a genuine domestic docket, with courts already sanctioning lawyers for AI-fabricated citations. For a jurisdiction whose AI-specific rules are policy, the IP posture and the international footprint are the profile's centre of gravity.
The site does three things: it maps every instrument that governs AI in South Africa, 16 of them, ordered by how hard each one binds; it reads the same instruments a second time by how squarely each is written for AI; and it records the matters where AI meets the law, with what was searched and found empty stated as plainly as what was found. Every claim traces to a primary source, and absences are recorded rather than papered over.
A membership claim enters this table only from an official list: a declaration annex, a treaty signature table, an organization's own record or an official government statement. South Africa's footprint is large, anchored by its 2025 G20 presidency and its BRICS membership.
Know of a ruling or proceeding in South Africa where AI is central? Send a source link; it will be reviewed before publishing.
Submit a case