The OSA’s impact on online memorialisation, collections, archives

Around two months ago, Nigel Nugawela delivered an excellent public lecture on archives, and archiving. Nigel’s a trained archivist, and some of what he talked about in April this year echoed what’s he covered in a speech delivered on International Archives Day held on 9 June 2023 at the National Archives, Colombo.

Nigel’s presentation in April touched on the Framework for National Archives and Records Management Legislation, which at the time was available for public comments (now closed). This piqued my interest, since the Online Safety Act (OSA) can be instrumentalised or interpreted in a way that sees all online collections of content, including archives, to be material that falls under it.

After I was subsequently sent the proposed legislation by Nigel, I studied it in relation to the OSA, and sent a detailed brief to him, and the head of the National Archives of Sri Lanka. The brief highlighted some of the incompatibilities in the proposed archives legislation, and the OSA. In the last section I noted,

The OSA places a significant, and sustained burden on the National Archives around compliance, and risk. Historically, the National Archives has been very poorly funded. Additional administrative, technical, human resource, legal, and curatorial burdens – from which there’s no escape given the overarching, and overbearing nature of the OSA – will likely not be a cost that’s covered through additional funding, leading to the possibility of significant lapses, gaps, and non or partial compliance issues. These in turn can be weaponised to shut down specific archives or render them inaccessible to anyone, and for as long as the OSA is in the country’s statute books.

Compliance with the Framework, and the Act is – based on what’s noted above alone – nightmarish, and near impossible to reconcile. It is unclear how even good faith arrangements to establish clear policies and guidelines to ensure compliance with both sets of laws can be meaningfully realised.

At a basic minimum, the National Archives will need to,

  1. Establish mechanisms to identify and flag online records that may be subject to OSA removal orders, and taking steps to proactively preserve copies of such records before they are removed.
  2. Engage with the Online Safety Commission and relevant stakeholders to explore alternative approaches to comply with OSA orders while minimising the impact on the integrity, and preservation of, as well as public (or at the very least, authorised) access to archives.
  3. Develop clear policies and guidelines for handling online records that may be subject to OSA removal orders, including protocols for preserving copies, documenting removal orders, and communicating with the public about the reasons for restricted access. For example, the National Archives could consider implementing a review process to identify and flag archives that may contain prohibited false statements. Access to such archives could be restricted, with clear explanations provided to the public about the reasons for the restrictions.

I understand the National Archives will seek guidance, and possible institutional exemptions from the Attorney-General’s Department around how best to negotiate the irreconcilably incompatible substance, and spirit of the OSA, with the proposed archives, and records management legislative framework.

Personal archives, and collections

While this may help the National Archives avoid the brunt of the OSA’s censorious design, it is of no help whatsoever to individuals who have already presented collections or archives online, and want to do so in the future. This hits home, and is a significant worry now because of what I’ve diligently done for well over two decades.

Some of the material I have collected, which now exist nowhere else on the Internet, web or social media, may well be construed to be violative of the OSA. Unlike the National Archives of Sri Lanka, I don’t have (and neither do I seek) a privileged pathway to the AG’s Department to iron out legal issues or seek indemnity. This means that those like myself are completely at the mercy of a draconian law that in the service of autocratic or censorious ends, can arbitrarily determine any collection of material online to be violative. This applies to online collections, and archives curated by anyone or any institution, anywhere in the world, and hosted on any online platform, product or service.

The cost of compliance

The second more general point I’ve made around all this is the cost of, and risks associated with compliance. The OSA introduces a risk framework to business-as-usual practices at institutions that is unprecedented in its complexity, extending from outward facing communications, content, and commentary through to even privileged internal, in-house, intranet, and cloud based exchanges, and material. The OSA can determine any IP address to fall under the law. This means that even an intranet hosted collection (i.e., not one that is publicly accessible over the Internet, but can be accessed from within a particular organisation’s network) can be reported by someone (with an axe to grind, for example) as potentially violating the OSA – in turn opening up the entire organisation’s infrastructure for scrutiny. This can be hugely disruptive, and extremely damaging.

Compliance can be costly too – because the OSA now requires everyone in the country (and actually, the world – but let’s leave this madness aside for now) – to do their own homework to ensure they aren’t in violative of a law that is impossible to be compliant with in the first instance. This Kafkaesque situation is brought about by broad language in the Act like “feelings of ill will” which can mean anything.

Whither (online) memorialisation?

In a country where acts of memorialisation offline continues to be marked by state-sponsored violent erasure, censorship, and pushback, online platforms serve as vital spaces for recording inconvenient truths, including eye-witness accounts, and other material from during the war, and after never recognised in mainstream accounts, official records or media.

This online memorialisation, and record-keeping, which necessarily involves the public presentation of digital collections, and potentially, archives of digitised material, are now at risk of going dark because of the OSA’s weaponisation. Curators, collectors, producers, editors, and owners of these collections, especially if in Sri Lanka now risk arbitrary incarceration, hefty fines or worse.

This can lead to very strange situations. For example, in 2024, Sri Lanka wanted to ban, and criminalise the distribution of kanji. As The Hindu noted,

To recall…desperate times, many Tamils prepare and serve kanji in the run up to May 18, marking the period as ‘Mullivaikkal remembrance week’, or as ‘Tamil Genocide Remembrance Day’ by some. Earlier this week, four Tamils were arrested in the Eastern Province while they were serving kanji, after police obtained a court order citing reasons of “public health” and revival of “terrorist activities”. The three women and a man were granted bail yesterday, amid sharp criticism of the government for “suppressing” Tamils’ right to memorialise.

I referred to this in a recent public lecture on the OSA, where I placed the word kanji against a slide that talked about the OSA’s massive blow to minorities, and memorialisation.

Under the Act, potentially any online presentation, capture or narrativisation of what by the state is criminalised, and banned offline, also risks being interpreted as violative of the law. Bearing witness to injustice online, and recording it for posterity, is now legally riskier under the OSA than it was during the war – which is quite extraordinary to even contemplate. This argument also applies to collections of content around 2022’s aragalaya, and go as far back as digitised records of material from during the war.

Though I hope the OSA’s life in Sri Lanka’s statute books is short-lived, what the National Archives of Sri Lanka may escape the worst of, nevertheless impacts everyone else who has some sort of an online collection or archive that can now be interpreted by anyone, at anytime, as violative.

These are just some of the significant, yet hidden impacts of Sri Lanka’s Online Safety Law. This is why I consistently maintain that the law must be rejected, and repealed.