Officials in New Delhi, who have had disagreements with the companies over restrictions on free speech, say the sites are not responding quickly enough to their requests to delete and trace the origins of doctored photos and incendiary posts aimed at people from northeastern India. After receiving threats online and on their phones, tens of thousands of students and migrants from the northeast have left cities like Bangalore, Pune and Chennai in the last week.
The government has blocked 245 Web pages since Friday, but still many sites are said to contain fabricated images of violence against Muslims in the northeast and in neighboring Myanmar meant to incite Muslims in cities like Bangalore and Mumbai to attack people from the northeast. India also restricted cellphone users to five text messages a day each for 15 days in an effort to limit the spread of rumors.
Officials from Google and industry associations said they were cooperating fully with the authorities. Some industry executives and analysts added that some requests had not been heeded because they were overly broad or violated internal policies and the rights of users.
The government, used to exerting significant control over media like newspapers, films and television, has in recent months been frustrated in its effort to extend similar and greater regulations to Web sites, most of which are based in the United States. Late last year, an Indian minister tried to get social media sites to prescreen content created by their users before it was posted. The companies refused and the attempt failed under withering public criticism.
While just 100 million of India's 1.2 billion people use the Internet regularly, the numbers are growing fast among people younger than 25, who make up about half the country's population. For instance, there were an estimated 46 million active Indian users on Facebook at the end of 2011, up 132 percent from a year earlier.
Sunil Abraham, an analyst who has closely followed India's battles with Internet companies, said last week's effort to tackle hate speech was justified but poorly managed. He said the first directive from the government was impractically broad, asking all Internet "intermediaries" - a category that includes small cybercafes, Internet service providers and companies like Google and Facebook - to disable all content that was "inflammatory, hateful and inciting violence."
"The Internet intermediaries are responding slowly because now they have to trawl through their networks and identify hate speech," said Mr. Abraham, executive director of the Center for Internet and Society, a research and advocacy group based in Bangalore. "The government acted appropriately, but without sufficient sophistication."
In the days since the first advisory went out on Aug. 17, government officials have asked companies to delete dozens of specific Web pages. Most of them have been blocked, but officials have not publicly identified them or specified the sites on which they were hosted. Ministers have blamed groups in Pakistan, a neighbor with which India has tense relations, for creating and uploading many of the hateful pages and doctored images.
A minister in the Indian government, Milind Deora, acknowledged that officials had received assistance from social media sites but said officials were hoping that the companies would move faster.
"There is a sense of importance and urgency, and that's why the government has taken these out-of-the-way decisions with regards to even curtailing communications," Mr. Deora, a junior minister of communications and information technology, said in a telephone interview. "And we are hoping for cooperation from the platforms and companies to help us as quickly as possible."
Indian officials have long been concerned about the power of modern communications to exacerbate strife and tension among the nation's many ethnic and religious groups. While communal violence has broadly declined in the last decade, in part because of faster economic growth, many grievances simmer under the surface. Most recently, fighting between the Bodo tribe and Muslims in the northeastern state of Assam has displaced about half a million people and, through text messages and online posts, affected thousands more across India.
Officials at social media companies, speaking on the condition of anonymity to avoid offending political leaders, said that they were moving as fast as they could but that policy makers must realize that the company officials have to follow their own internal procedures before deleting content and revealing information like the Internet protocol addresses of users.
"Content intended to incite violence, such as hate speech, is prohibited on Google products where we host content, including YouTube, Google Plus and Blogger," Google said in a statement. "We act quickly to remove such material flagged by our users. We also comply with valid legal requests from authorities wherever possible."
Facebook said in a statement that it also restricts hate speech and "direct calls for violence" and added that it was "working through" requests to remove content. Twitter declined to comment on the Indian government's request.
Telecommunications company executives criticised the government's response to the crisis as being excessive and clumsy. There was no need to limit text messages to just five a day across the country when problems were concentrated in a handful of big cities, said Rajan Mathews, director general of the Cellular Operators Association of India.
"It could have been handled much more tactically," he said.
Others said the government could have been more effective had it quickly countered hateful and threatening speech by sending out its own messages, which it was slow to do when migrants from the northeast began leaving Bangalore on Aug. 15.
"It has to also reach out on social networking and Internet platforms and dismantle these rumors," Mr. Abraham said, "and demonstrate that they are false."