Meta Platforms has updated Instagram’s content recommendation systems in India to reduce the visibility of sexually suggestive material and violent content, introducing more advanced artificial intelligence models to identify potentially objectionable posts before they are recommended to users. The changes are designed to strengthen the platform’s default protections and address the types of content Indian users have said they do not want to encounter while browsing the app.
The updated measures apply to content recommended from accounts that users do not follow, rather than introducing a blanket ban on such material. The changes cover Instagram’s Reels, Explore and Feed recommendations. Meta said the improvements have helped reduce the frequency with which users report recommended content by approximately 50% compared with a year ago. The announcement comes amid increasing scrutiny of social media algorithms and their role in determining what content reaches large audiences in India.
Instagram Uses Advanced AI to Filter Sensitive Content
Meta said it has strengthened Instagram’s recommendation systems over the past year, with a particular focus on identifying content that does not meet its Recommendations Guidelines.
The company is using more capable AI models to assess content directly against these guidelines. Previously, platforms often relied heavily on user reports and feedback to identify material that should not be recommended. The upgraded systems are designed to detect potentially objectionable content proactively, before users need to report or hide it.
The measures specifically target sexually suggestive material and content depicting violence.
According to Meta, the improvements allow Instagram to prevent more content that falls outside its recommendation standards from appearing in users’ discovery experiences.
The company said the frequency of user reports about recommended content has fallen by half compared with the previous year. However, Meta has not disclosed the total volume of posts affected, the percentage of recommendations filtered or the detailed methodology behind the reported reduction.
This means the 50% figure refers to the frequency of user reports, not a confirmed 50% reduction in all sexually suggestive or violent content on the platform.
Changes Apply to Instagram Reels, Explore and Feed
The India-specific updates cover three major areas of Instagram where users regularly discover content from accounts they do not follow.
| Instagram feature | How recommendations work |
|---|---|
| Reels | Suggests short-form videos based on user interests and activity |
| Explore | Displays photos, videos and accounts beyond a user’s existing following list |
| Feed | Includes recommended posts alongside content from followed accounts |
These recommendation systems play an important role in determining which creators and posts receive visibility beyond their existing audiences.
For example, a user watching fitness videos may receive recommendations for workout creators they have never followed. Similarly, someone interested in entertainment may see suggested videos from unfamiliar accounts.
Meta’s latest changes aim to reduce the chances that potentially sensitive or objectionable material appears through these recommendations.
The distinction is important for creators and users. The company is changing what its algorithms promote, not necessarily removing every post that contains sensitive material.
Recommendation Guidelines Are Different From Content Removal Rules
Instagram operates under two separate sets of standards that determine how content is treated on the platform.
Community Standards: These rules identify content that is prohibited on Instagram. Posts that violate these standards can be removed from the platform.
Recommendations Guidelines: These rules determine whether content can be promoted to people who do not already follow the account that published it.
Some posts may remain accessible on Instagram while being excluded from recommendations because they contain material considered sensitive, objectionable or unsuitable for wider distribution.
For example, content that does not violate Instagram’s Community Standards may still be considered inappropriate for recommendation to a broad audience.
The distinction gives Meta an additional way to manage content distribution without treating every sensitive post as a violation requiring removal.
The company also says that content considered unlawful in a particular country is restricted in that market.
Government Scrutiny Adds Pressure on Social Media Algorithms
The announcement comes as Indian authorities increase their scrutiny of how social media platforms rank and recommend online content.
On October 10, 2026, the Ministry of Electronics and Information Technology (MeitY) issued an advisory asking Meta to review its ranking and recommendation systems.
The government raised concerns about potentially misleading, manipulated or out-of-context material receiving disproportionate visibility through platform algorithms.
The advisory focused on the way recommendations can amplify content relating to public events and developments, potentially contributing to confusion or public-order concerns when information lacks adequate context.
Meta has not explicitly confirmed that its latest Instagram changes were introduced in response to the government’s advisory.
Nevertheless, the timing places the announcement within a broader discussion about the responsibilities of social media companies in managing algorithm-driven distribution.
Recommendation algorithms influence what millions of users discover every day. Their decisions can affect the reach of individual creators, the spread of information and the visibility of potentially harmful material.
Instagram Gives Users More Control Over Recommended Content
Alongside its default recommendation protections, Instagram offers several tools that allow users to customise what they see.
These features are particularly relevant because automated systems cannot always understand every user’s individual preferences.
Some of the available controls include:
| Feature | What it allows users to do |
|---|---|
| Sensitive Content Control | Choose to see less sensitive material in recommendations |
| Political Content Control | Reduce political content recommended from accounts users do not follow |
| Not Interested | Tell Instagram that a particular post or recommendation is unwanted |
| Hidden Words | Filter selected words, phrases and emojis from comments and message requests |
| Block | Prevent selected accounts from finding a profile, viewing posts or sending messages |
The Sensitive Content Control feature is available through Instagram’s Settings and Activity section under Content Preferences. It allows users to limit sensitive material across Search, Reels, Explore, suggested accounts and in-Feed recommendations.
Meta said its systems in India already exclude the most mature content from recommendations. Users can use the control to further reduce exposure to other sensitive material.
The Not Interested option also plays a role in improving recommendations. When users select it, Instagram receives feedback that a particular post or type of content is not relevant to them.
Meta says this feedback helps its systems understand what people prefer not to see.
Teen Accounts Receive Additional Safety Protections
Instagram has also introduced additional safeguards for younger users through its Teen Accounts feature.
Teen Accounts were introduced in India in 2025 and automatically place teenage users in a 13-plus content setting governed by Instagram’s age-appropriate content guidelines.
These protections are intended to reduce teenagers’ exposure to material that may be inappropriate for their age group.
The latest recommendation changes add to the platform’s wider efforts to manage sensitive content and improve the experience of younger users.
The issue is particularly significant because teenagers can encounter recommended videos and posts from creators they do not follow, making algorithmic controls an important part of online safety.
However, recommendation filters are only one part of the broader challenge. User behaviour, account settings, content moderation and parental involvement also influence the online experiences of younger audiences.
What the Changes Mean for Instagram Creators and Businesses
The updated recommendation systems could also affect creators, influencers and businesses that depend on Instagram to reach new audiences.
Content from accounts users do not follow is an important source of discovery for creators. Reels and Explore recommendations can help smaller accounts attract viewers without relying entirely on their existing follower base.
However, stricter recommendation filters may reduce the distribution of material that Instagram considers unsuitable for wider promotion.
Creators may need to pay closer attention to the platform’s Recommendations Guidelines, particularly when publishing content involving sensitive themes, suggestive imagery or depictions of violence.
For businesses, the changes highlight the importance of creating content that is suitable for broad distribution rather than focusing exclusively on engagement.
Meta has not announced any specific changes to advertising eligibility or provided estimates of how the updated systems could affect creator reach and business accounts.
The Bigger Picture
Instagram’s India-specific changes reflect the growing importance of algorithmic responsibility in the social media industry. Platforms no longer simply host content uploaded by users; their recommendation systems actively determine which posts receive additional visibility. By using more advanced AI models to identify sensitive material, Meta is attempting to strengthen safeguards before potentially objectionable content reaches wider audiences. The move also comes at a time when Indian regulators are paying closer attention to how algorithms amplify information.
For creators, advertisers and users, the distinction between removing content and restricting its recommendation will remain important. Posts that comply with Instagram’s Community Standards may still lose access to wider discovery if they do not meet recommendation requirements. While Meta reports a 50% reduction in user reports about recommended content, the company has not disclosed enough underlying data to independently assess the overall impact of the changes.
Looking Ahead
Meta says it will continue refining Instagram’s recommendation systems as technology and user behaviour evolve. The effectiveness of the India-specific measures will depend on how accurately its AI models identify sensitive material, how consistently the rules are applied and whether users notice a meaningful improvement in their recommendations. Further transparency about the volume of content filtered and the methodology behind its reported results could help users and regulators evaluate the changes.
For India’s social media market, the development adds to a wider debate about the responsibilities of platforms that use algorithms to distribute information at scale. Instagram will need to balance user safety, creator visibility and personalised discovery while responding to regulatory expectations. The long-term impact will depend not only on stronger AI detection but also on transparent enforcement and meaningful user controls.
Get the day’s top stories in your inbox
One concise email. No spam, unsubscribe anytime.



