Instagram is a very popular social media app based on sharing photos and videos, publicly and to selected users as well as via direct, private message. It was launched in 2010 and since April 2012 has been owned by Facebook, another giant in the social media industry. In less than the decade of its existence, Instagram has grown a very large and active community, where users can interact with their friends and “followers” as well as other communities who maintain a presence there, public figures, media sources, and corporate brands.

All of these wildly different groups, from all over the world, sharing content and commentary on one platform, is exciting and promises many opportunities for collaboration. Along with these positive connections, though, of course come negative surprises and possibilities for challenges and abuses. With all the influence Instagram has through its popularity comes also responsibility for defining the standards and limitations of the community as well as what it will put out into the internet and the world.

Instagram has faced its share of criticism for its efforts to implement and maintain effective controls and reporting mechanisms.   Instagram relies heavily on user reporting of inappropriate content, such as posts depicting illegal activity or the use of “coded” hashtags and emojis to conceal but continue on with such practices. Understandably, even the most aggressive attempts to keep up with the pace of this behavior on social media will fall behind quickly, leading to criticism the community is unsafe. When Instagram is too proactive or reaches in deleting comments, posts, or users, however, then controversy about overreaching into privacy and expression begins in response.

Kevin Systrom, one of the original creators of Instagram and its current CEO, wants to work this balance between protection from abuse and freedom of expression. Under his leadership, Instagram is dedicated to ensuring that the content and tone on the platform is compliant with its community guidelines. Changes to the comments sections on photos – including allowing users to filter out comments that had certain words, or to post photos without comment sections available – are intended to encourage safer self-expression by the posters who might otherwise fear harassment or offensive content in response below their photos.

Platforms such as Instagram, of course, can never be neutral – any technology’s relationship with its user is one that is fraught with moral concerns, starting right at the ethics of its design, which is made only more complex by algorithms, robot users, and the real users who make their own decisions about the content to share and promote that run the gamut from universally appropriate to offensive, harassing, or even illegal. In such a context, applying a code of ethics is a very hard task, but perhaps it is the inherent difficulty of doing this that makes it so important to try.

Creating filters and tools to hide and promote, prevent and engage, either when deployed by the community management behind the scenes or when elected by users, is just the beginning of the design choices engineers have made at Instagram to implement technical responses to problematic tone in some corners of the platform. Instagram tries to deploy artificial intelligence to help also, to sort real posts from fake and to learn from the data to understand why innocent comments or content may be abusive to the context, a concept called word embeddings. AI has its limitations, of course, but in any rules-based approach to governance it’s necessary to start with something good and then make continual efforts to make it better, rather than leave risks un-addressed while in hopeful pursuit of the best.

Time will tell how effective Instagram’s efforts to make the platform a safer place for expression really are, and what they really accomplish – a place which is open for creative sharing and communication creation, but not to toxicity and abuse, or a censored, sanitized, disingenuous photo collection where self-expression is restricted and speech censored? Perhaps Instagram will succeed in going against the tide on the internet and in much of life, where the level of social discourse seems to have gone low, tinged by anger and dark with people’s worst impulses, and make a place where the conversation can be a bit more civil, even if it has to be filtered first to get there.

For more detail on Kevin Systrom’s ambition of making Instagram a safe haven and role model platform on the internet, and the challenges that both motivate and complicate this mission, see Nicholas Thompson’s story on Wired.