These subversive activities have bedevilled the political arena for over a decade, doing serious damage to the democratic system and the legitimacy of the intelligence services.

Interference in party politics by intelligence officers persists mainly because the executive is complicit in the mischief and has therefore not cracked down on it. It also persists because the legislative mandate of the intelligence services allows them to meddle in politics – it even encourages them to do so.

The problem of inappropriate intelligence mandates is evident in the General Intelligence Laws Amendment Bill (Gilab), currently under review by a parliamentary committee. The bill provides for the establishment of the new State Security Agency (SSA), which is an amalgamation of the National Intelligence Agency (NIA) and other civilian intelligence bodies.

Gilab provides that the SSA must gather and analyse domestic intelligence with the aim of identifying any threat or potential threat to the security of the Republic or its people. The bill defines “domestic intelligence” to mean “intelligence on any internal activity, factor or development which is detrimental to the national stability of the Republic, as well as threats or potential threats to the constitutional order of the Republic and the safety and well-being of its people”.

These provisions are identical to those that governed the NIA. The continuity is surprising since it is well known that the NIA’s mandate was way too broad, vague and open to interpretation. The terms “security of the Republic and its people”, “national stability” and “threats to the constitutional order”’ are notoriously imprecise and drew the agency into the field of politics.

Between 1994 and 2008, NIA reinterpreted its mandate three times without the results of this process being revealed to Parliament and the public. At one stage the mandate was construed so expansively as to encompass the focus of virtually every state department. According to an internal directive issued in 2003, the agency was expected to “inform decision-makers about every aspect of human endeavour upon which good order and the prospects for a prosperous future depend”. This was patently impractical, if not absurd.

The broad legislative mandate also led to NIA adopting what it called a “political intelligence function”. This entailed monitoring and reporting on transformation within government department, on competition within and between political parties and on the impact of the government’s policy decisions.

An intelligence orientation of this kind is outrageous in a democratic society.

NIA itself became unhappy about its politicised and sweeping mandate. In 2008 it told the Ministerial Review Commission on Intelligence that the agency should have a narrow mandate, that the mandate should concentrate on serious crimes and that the political intelligence function should be abandoned.

The commission, set up by former intelligence minister Ronnie Kasrils, agreed with NIA. It proposed that the intelligence legislation should emulate the Canadian and Australian models that exclude lawful political activity from the remit of the domestic intelligence agency.

Gilab only takes partial account of this proposal. It states that the SSA’s responsibility to protect “national security” covers terrorism, espionage and subversion but excludes “lawful political activity, advocacy, protest or dissent”.

This is a positive step but the bill does not amend the domestic intelligence mandate or the definition of “domestic intelligence”. It therefore fails to address adequately the danger of intelligence interference in politics.

This danger is heightened by Gilab’s provisions on the counter-intelligence mandate of the SSA. Here the bill ignores completely the concerns of NIA and the Ministerial Review Commission.

The bill states that the SSA shall conduct and co-ordinate counter-intelligence. The definition of “counter-intelligence” includes “measures and activities conducted, instituted or taken to impede and to neutralise the effectiveness of foreign or hostile intelligence operations” and “to counter subversion, sedition, treason and terrorist and related activities”.

In its 2008 submission to the Ministerial Review Commission, NIA argued that a similar formulation in the National Strategic Intelligence Act did not provide clear guidelines on counter-intelligence. What is meant by “impede”, “neutralise” and “counter”? Which measures are legitimate and which are illegitimate? Gilab leaves these questions unanswered.

The result is that offensive counter measures, which carry the risk of infringing constitutional rights and interfering in lawful political activity, are not subject to proper legal definition and constraints. The consequent risk is that these measures will violate rights without sufficient cause, safeguards and sense of caution.

The White Paper on Intelligence of 1994 contains an excellent policy prescription in this regard: no intelligence organisation “shall be allowed to carry out any operations or activities that are intended to undermine, promote or influence any South African political party or organisation at the expense of another by means of any acts (eg ‘active measures’ or ‘covert action’) or by means of disinformation”. This position should be included in Gilab.

Parliament now has an opportunity to strengthen the intelligence legislation and reduce the potential for mischief by the spies and their political allies. If it spurns this opportunity, it will have failed to protect the constitutional order against a subversive tendency that emanates from within the state itself.

This article appeared in the Cape Argus of 25 April 2012